(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Proceedings of the American Society of Agronomy"

OAK ST HDSF 



THE UNIVERSITY 
OF ILLINOIS 
LIBRARY 

65Q6 
AMR 

v.l0cop3 



lIBfMHr 



This book has been DIGITIZED 

and is available ONLINt. 



Digitized by the Internet Archive 
in 2013 



http://archive.org/details/proceedingsofame1019amer 



JOURNAL 

OF THE 

AMERICAN SOCIETY 
OF AGRONOMY 



VOLUME 10 



1918 



PUBLISHED BY THE SOCIETY 



PRESS OF 
THE NEW ERA PRINTING COMPANY 
LANCASTER, PA. 



DATES OF ISSUE. 



Pages 1-48, January 20, 1918. 
Pages 49-96, February 15, 1918. 
Pages 97-144, March 20, 1918. 
Pages 145-192, April 25, 1918. 
Pages 193-224, June 15, 191 8. 
Pages 225-264, September 21, 1918. 
Pages 265-312, Xovember 1, 1918. 
Pages 313-360. February 8, 1919. 



iii 



ERRATA. 



Pa.ye 25. Table 2, column headings "Hard" and "Soft" are reversed 

throughout'. 

Page 83. name of author of paper is Albrecht, not Aldrecht. 
Date of issue of No. 4 should be April 25, 1918. 
Plate 4. facing page 151, figure 2 is reversed. 

Page 180, line 7 was carried down with footnote 2 and second line of this 
footnote precedes the first. 

Pa: 1 -'So, line 8, and page 281. line 1. read " Plate 0" for " Plate 7." 



Iv 



Am R 



CONTENTS. 
No. i. JANUARY. 

Page. 

Stevens, O. A. — Variations in Seed Tests Resulting- from Errors in 

Sampling (Figs. 1-3) I 

Dunnewald, T. J. — Vegetation as an Indicator of the Fertility of Sandy- 
Pine Plains Soils in Northern Wisconsin (Fig. 4) 10 

Freeman, Geo. F. — A Mechanical Explanation of Progressive Changes in 

the Proportions of Hard and Soft Kernels in Wheat 23 

MacIntire, W. H. — The Growth of Sheep Sorrel in Calcareous and Dolo- 

mitic Media (PI. 1) 29 

Walworth, E. H., and Smith, L. H. — Variations in the Development of 

Secondary Rootlets in Cereals 32 

Call, L. E., and Sewell, M. C. — The Relation of Weed Growth to Nitric 

Nitrogen Accumulation in the Soil 35 

Hartwell, B. L., and Pember, F. R. — Aluminum as a Factor Influencing 

the Effect of Acid Soils on Different Crops 45 

Agronomic Affairs. 

Membership Changes 47 

Notes and News 48 

No. 2. FEBRUARY. 

Waller, Adolph E. — Crop Centers of the United States (Figs. 5-12) 49 

Albrecht, Wm. A. — Changes in the Nitrogen Content of Stored Soils 83 

Ball, Carleton R., and Clark, J. Allen — Naming Wheat Varieties 89 

Agronomic Affairs. 

Annual Dues for 1918 94 

The Society's Honor Roll 95 

Membership Changes 9.S 

Notes and News 96 

No. 3. MARCH. 

Bizzell, J. A., and Lyon, T. L. — The Effect of Certain Factors on the 

Carbon-Dioxide Content of Soil Air (Figs. 13-21) 97 

Snyder, Harry — Wheat Breeding Ideals 113 

Hayes, H. K. — Natural Cross-Pollination in Wheat 120 

Hayes, H. K. — Normal Self-Fertilization in Corn 123 

McCall, A. G., and Richards, P. E. — Mineral Food Requirements of the 
Wheat Plant at Different Stages of Its Development (Pis. 2 and 3 

and Figs. 22 and 23) 127 

Garber, R. J., and Arny, A. C— Relation of Size of Sample to Kernel- 
Percentage Determinations in Oats 134 

v 

537C43 



vi 



CONTENTS. 



Page. 



Agronomic Affairs. 

Membership Changes 143 

Notes and News 144 

No. 4. APRIL. 

Love, H. H., and Craig, YY. T— Methods Used and Results Obtained in 

Cereal Investigations at the Cornell Station (PI. 4 and Fig. 24) 145 

Emerson, Paul — A Simple Method of Demonstrating the Action of Lime 

in Soils 158 

McKee. Roland — Glandular Pubescence in Various Medicago Species 150 

Piper. C. V. — Cutthroat Grass (Panicum combsii) 162 

Hill, C. E. — A Drill for Seeding Nursery Rows 165 

Spragg. F. A. — Red Rock Wheat and Rosen Rye 167 

Montgomery. E. G. — The Identification of Varieties of Oats in New York 171 
Coe, H. S. — Origin of the Georgia and Alabama Varieties of Velvet Bean 

(Figs. 25 and 26) 175 

Karraker, P. E. — The Value of Blue Litmus Paper from Different Sources 

as a Test for Soil Acidity 180 

Biggar, H. Howard — Primitive Methods of Maize Seed Preparation 183 

McClelland, C. K.— The Time at Which Cotton Uses the Most Moisture 185 
Agronomic Affairs. 

Official Changes 189 

Annual Dues for 1918 189 

Membership Changes 190 

Roll of Honor 191 

Notes and News 191 

No. 5. MAY. 

Le Clerc, J. A., and Davidson, J.— The Effect of Sodium Nitrate Applied 
at Different Stages of Growth on Yield, Composition, and Quality 
of Wheat-2 193 

Davisson, P>. S., and Sivaslian, G. K. — The Determination of Moisture 

in Soils 198 

BotH NAKiAN, Sarkis— The Mechanical Factors Determining the Shape of 

the* Wheat Kernel (Fig. 27) 205 

BOLTS, G EL— Loss of Organic Matter in Clover Returned to the Soil 210 

Le Clerc, J. A., Bailey, L. H., and Wesslino, IT. L.— Milling and Bak- 
ing Tests of Einkorn, Emmer, Spelt, and Polish Wheat 215 

' I F. -Comparative Smut Resistance of Washington Wheats.... 218 

Agronomic Affairs. 

Membership Changes 222 

Roll of Honor 223 

Notes and News 223 

No. 6. SKI'TKMI'.KK. 

ROUURMOt OSWALD, and Skfnnf.r, J. J.- The Triangle System for Fer- 
tilizer Kxprrimcnti (]*]%. 5-7 and Figs. jK-41) 225 



CONTENTS. vii 

Page. 

Hendry, G. W. — Relative Effects of Sodium Chloride on the Development 

of Certain Legumes 246 

Hutcheson, T. B., and Wolfe, T. K— Relation between Yield and Ear 

Characters in Corn 250 

Leonard, Lewis T., and Turner, C. F. — Influence of Ceretoma trifurcata 

on the Nitrogen-Gathering Functions of the Cowpea (PI. 8) 256 

Agronomic Affairs. 

Notice of Annual Meeting 262 

Membership Changes 262 

Roll of Honor 263 

Notes and News 263 

Nos. 7-8. OCTOBER-NOVEMBER. 

Alway, F. J., McDole, G. R., and Trumbull, R. S. — Interpretation of 

Field Observations on the Moistness of the Subsoil 265 

Bryan, W. C. — Hastening the Germination of Bermuda Grass Seed by the 

Sulfuric Acid Treatment (PI. 9) 279 

Merkle, F. G. — The Decomposition of Organic Matter in Soils (Figs. 

41-46) • • 281 

Cowgill, H. B. — Cross-Pollination of Sugar Cane 302 

Agronomic Affairs. 

Annual Meeting in Baltimore 307 

Error in the January Number 307 

Membership Changes 307 

Roll of Honor 308 

Notes and News 309 

No. 9. DECEMBER. 

Lyon, T. Lyttleton. — Influence of Higher Plants on Bacterial Activities 

in Soils (Presidential Address) 313 

Warburton, C. W. — The Preparation of Manuscripts for Publication 322 

Agronomic Affairs. 

Delay in Publication of the Journal 326 

The Year's Work 326 

Honor Roll 325 

Notes and News 3 2 7 

Report of the Secretary-Treasurer 33^ 

Financial Statement 330 

Address List of Members 33 2 

Lapsed Members 342 

Minutes of the Eleventh Annual Meeting 343 

Report of Committee on Standardization of Field Experiments 344 

Report of the Committee on Terminology 354 

Report of the Editor 354 

Index ' 356 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. io. January, 1918. No. r. 

VARIATIONS IN SEED TESTS RESULTING FROM ERRORS IN 

SAMPLING. 1 

O. A. Stevens. 
Introduction. 

Little attention appears to have been given to' careful study of the 
accuracy of seed testing. Some experiments have been made during 
the past few years by sending to different laboratories samples drawn 
from a common bulk. The results of one of these has been pub- 
lished (Stone, 1913), 2 but so far as the writer could find from avail- 
able literature no attempt had been made to analyze the fundamental 
factors to' which variations are due. In this paper the results are 
presented of some investigations conducted during 1914 and 191 5 
when routine duties were not pressing. 3 

Causes for Variation. 

Variations in the results of seed testing are fundamentally of two 
kinds, and are to a great extent impossible to avoid. One is purely 
mathematical, the other personal or to a large extent economic. The 
direct causes may be listed as follows : 

1 Contribution from the Department of Botany, North Dakota Agricultural 
College. Received for publication Ala}' 14, 1917. 

2 Bibliographical citations in parentheses refer to " Literature cited," p. 19. 

3 The author is indebted to Prof. I. W. Smith for assistance with the mathe- 
matical portion of the work. Credit is also due student assistants as follows : 
Mr. M. S. Hagen for all of the germination work except two lots of 50 
each; Mr. A. M. Christensen for the series of alfalfa purity tests and two of 
germination; and Mr. Sidney Hooper for one series of purity tests of flax. 

I 



2 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



In Purity Tests. 



In Germination Tests. 



Imperfect mixing. 

Effect of random sampling. 

Errors in weighing. 

Effect of personal selection. 

Errors of identification. 



Imperfect mixing. 

Effect of random sampling. 

Errors in counting. 

Effect of personal selection. 

Improper conditions for germination. 

Special conditions of the seed. 



Considering a sample from its point of origin the first source of 
variation is in drawing a small sample from a large bulk. The ac- 
curacy with which this sample represents the bulk depends upon the 
uniformity of the latter. This would depend upon the uniformity 
of the field or whether different lots had been mixed to make up the 
stock. When samples are drawn and mailed by various parties, there 
is greater possibility of variation than if all could be taken by the 
same person. When the analyst receives the sample and takes from 
it a standard sample for analysis the same factors are again operative. 

Before the standard laboratory sample is taken, the entire lot 
should be mixed as thoroughly as possible. If the seed has been well 
cleaned, it is doubtful if any mixing is necessary. In poorly cleaned 
seed there are usually elements of different nature which tend to col- 
lect at certain places, such as sand, fine dirt, or small seeds at the 
bottom, and broken stems, etc., at the top. In such samples it may 
be difficult to make an even mixture, and it is desirable to separate 
such materials roughly from the entire sample. It may be observed 
that in such cases the error in drawing the sample from the bulk is 
likely to be large, and a careful analysis is scarcely worth the while. 

After a sample has been mixed as well as possible, a small por- 
tion will not accurately represent the entire amount. If, for example, 
wt take one seed out of a lot of one hundred which are half of one 
kind and half of another, it will give no estimate of the proportions 
of the mixture. As we take a larger number, the degree of accuracy 
increases but can not be perfect unless the entire number is used. 
This is what is here referred to as fundamentally a source of mathe- 
matical error. Tin's error ran not be avoided, but it is possible to 
determine it- approximate limits under given conditions and thus 
make allowance for it. 

There is a certain amount of persona] error in weighing out a 
standard sample for analysis, or in counting a certain number of 
eedl for genninatiofl tests, but this is atl occasional factor rather 
than I COD tanl One, U arc mOSl Of those under consideration. 

The error Of personal Selection is of considerable importance. In 
germination work it may be a tendency to pick out the better seeds. 



STEVENS : VARIATIONS IN SEED TESTS. 



3 



There should, of course, be a standard method which would prevent 
this, yet it is not entirely possible. The seeds for germination should 
be selected from the pure seed after the purity test has been com- 
pleted ; but the length of the germination period is the factor which 
determines the time necessary for a report. Under the usual rela- 
tion of amount of work to the equipment of the laboratory, time is 
saved by starting a germination test at once. 

There is in nearly every sort of seed an indefinable line of separa- 
tion between pure seed and inert matter, i. e., shrunken seeds, broken 
seeds, and grass flo'rets without caryopses. One may make an arbi- 
trary division, but there will be cases coming so close to the line that 
they might be placed on either side. For instance, if half a seed be 
placed with the inert matter, and more than half with the pure seed, 
there will be found individuals that are not readily placed. A single 
instance of this in case of flax will make a difference of 0.02 percent 
of 10 grams. With wheat a single such grain would amount to 0.05 
percent of 30 grams. 

With grasses we may consider florets without caryopses as inert, 
but they occur in all degrees of development and are not always 
easily placed in either one or the other class. It should be noted, 
however, that these things occur most frequently in low-grade seed, 
in which small errors are not of great consequence. 

The error of identification should not be of consequence in most 
cases, but there are instances where it is likely to be. We may simply 
mention, in passing, the difficulty of separating such seeds as alfalfa 
and sweet clover, meadow fescue and perennial ryegrass, species of 
bluegrass, species of Agropyron, etc. 

The variations so far discussed apply to both purity and germina- 
tion tests. There are also causes of variation peculiar to the latter, 
such as metho'ds of testing and various factors modifying the condi- 
tion of the seed. Certainly, unusual variations may be expected if the 
temperature and moisture conditions of the germination chamber are 
not adapted to the seed. The failure to secure satisfactory results 
by handling bluegrass or other small and slowly germinating seeds 
by the same methods commonly used for clovers or cereals is a good 
example of this. Seeds not sufficiently aged for prompt germina- 
tion may respond better to conditions somewhat different from those 
usually employed for that kind of seed. Fresh seed o'f lettuce is one 
example which has caused considerable trouble in this way. Barley 
and oats slightly damaged by exposure to water or frost have also 
been a source of difficulty. 



4 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



By the economic factor is meant the lack of funds to secure proper 
equipment or to employ workers of sufficient ability and training. 

If the results of a series of analyses of samples drawn from a 
carefully mixed bulk and submitted to a number of laboratories are 
examined (Stone, 191 3), the effect of these two primary factors of 
error is readily seen. The mathematical variation may be readily 
demonstrated from some of the germination tables, as will be shown 
later in connection with the experimental data presented. From an 
inspection of the purity table of bluegrass .or orchard grass it may 
be seen that some reports give an unusually high percentage of pure 
seed and a correspondingly low percentage of inert matter. This is 
no doubt due to the inclusion in pure seed of the florets containing 
no caryopses. This may have been due to insufficiently trained 
analysts or to lack of equipment (vertical air blast separator, etc., 
without which such separation is quite difficult). From the table of 
bluegrass germination we note some very low results and some blanks 
probably due to a lack of equipment. 

Much of the personal factor depends upon experience, not only 
general, but with an individual class of seeds. The workers of a 
laboratory in a State where a certain crop is grown little or not at all 
can not be expected to be as familiar with the seeds of that crop as 
those who have constant occasion to examine them. 

Experiments with Germination Tksts. 

The work here described was undertaken to determine the mathe- 
matical error in germination tests. The general plan was to make a 
series of 50 tests from one lot of seed and to calculate the standard 
ilcviati' n and probable error, these tests being made simultaneously 
by one person. 

To reduce the sources Of variation as much as possible, it seemed 
d< sirable to make a scries of theoretical tests. A sample of seed to 
l>c t< «t«-d for germination ma\ he regarded as a mixture of two sorts, 
the "lie triable, the other not. The result then depends upon the per- 
fection of the mixture and the chance of selection in a given number. 

Alter some preliminary experiments a white-seeded kafir was 
Selected as besl suited to tlx- purpose. I 'art of the seed was stained 
with Delafield'j hematoxylin, the alcoholic solution being used in 
order to prevent, as far as possible, any change in the seed. 

The first question was to determine what mixing was necessary. 
For thi> purpose grains of each color were placed together and 
poured from one dish into another. Small pans holding several times 



STEVENS: VARIATIONS IN SEED TESTS. 5 

this amount were used and the seed poured from one to the other at 
the rate of fifty times per minute, the pans being about 6 inches apart. 
This process was repeated 25, 75, and 200 times and then the series 
repeated with the results shown in Table 1. 



Table i. — Probable error in selecting 100 seeds from a mixture of equal parts 

of two sorts. 



Number of trials. 


Mixed 25 times. 


Mixed 75 times. 


Mixed 200 times. 


First 

Second 


3-IO 

••! 3-OS 


2.88 
2.83 


2.83 
2.82 



The results of the trials are very close, in fact remarkably so, con- 
sidering the differences in later experiments. It is rather surprising 
also how little difference appears from the longer continued mixing. 
In the following work 75 mixings were made, the operation being 
as nearly uniform as possible throughout the series. In counting- 
out 100 seeds the practice followed was to pour the entire lot upon 
the work table and, by placing a hand on each side, draw out one end 




99 9J- ?0 30 70 Go SO 



Fig. 1. Graph showing probable error in selecting 100 seeds from mixtures 
of different percentages. G, Percentage of germination ; E, percentage of prob- 
able error ; a, no hard seed present ; b, 5 percent dead, the remainder hard ; c, 
10 percent dead; d, from Table 12 (Rodewald, 1889, p. no). 

of the pile to a narrow point. It would be preferable to have a 
mechanical means, such as weighing out the number. This was not 
used in the present case on account of the impossibility of taking 
just 100 seeds and the work involved in calculating odd numbers. 



6 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



The method used should obviate the possibility of unconscious selec- 
tion. 

The second problem was to determine the error for a given per- 
centage. In a 100 percent stock there could be no variation. As. the 
percentage decreases from this point we may expect the variation to 
increase until 50 percent is reached, then decreasing again to zero. 
In actual tests the poorly germinating seed might show a compara- 
tively greater variation on account of failure to respond to slightly 
varying conditions. 

A series of mixtures was prepared and carried through with the 
results shown in Table 2. Inasmuch as the hard seeds of legumes 
introduce a third factor, two additional series were run, using 5 and 
10 percent respectively as dead, the remainder being considered hard 
(dead represented by seeds colored with an alcoholic solution of 
erythrosin). 



Table 2. — Probable error in selecting 100 seeds from mixtures of different 

percentages. 



Percentage 
calculated 
viable. 


Error in viable series. 


Error in hard series. 


Error in dead series. 


With 0% 
hard. 


With 5%. 
dead. 


With 10% 
dead. 


With 5% 
dead. 


With 10% 
dead. 


5%. ■ 


10%. 


50 


2.88 


3 A* 


3-67 


3.60 


3-14 


I.92 


1.98 


60 


2.89 


3-42 


3-29 


3-44 


3-H 


I.84 


2.27 


70 


2.91 


324 


3-13 


308 


2.98 


i-55 


1.94 


80 


2.69 


2.95 


2.60 


2.38 


2.07 


1. 14 


1.98 


QO 


2.22 


i-95 




1.36 




1.28 




95 


1.30 














97 


115 














99 


.64 















I he results in the first three columns are shown graphically in 
figure 1. In the results of tin: second and third lines of the second 
column an extreme variation was obtained (2.81 and 3.99). A 
second trial of tin- was made with the results shown. 

\\ here onl) two types of seeds are present the variation of the one 
mn~t be tlx- same as that of the other, since the second always equals 
IOO minus the first. Therefore, the variation in a 10 percent series 
: be the same, at least theoretically, as in a 90 percent series, 
nd the greatest variation would he found when each is 50 per- 
cent. When three typo are present the number of possible combina- 
tions is increased and the variation of one scries need not he the 
-iiim- • either of the others. The greatest variation would he ex- 
pected when the percentage of each is equal. Accordingly, one lot 
Wai made up in this manner and the probable error found to be 3.83, 



STEVENS : VARIATIONS IN SEED TESTS. 



7 



3.92 and 3.01 respectively for the three sorts. The deviation of each 
is shown in figure 2. 

Two more series were run with this material and will be mentioned 
later. Several series of germination tests were also made and the 
probable error calculated in the same manner. In order to eliminate 
other factors as far as possible, seeds were used which do not offer 
special difficulties. All tests were made between folds of blotting 



/ 

1 


- 




























1 

1 

—h 


& 

\ 




























1 

1 


\ 








1 




















1 1 
1 / 


I — < — 
\ 








T 




















— *— *- 










1 




















' / / 




"v \ 
\ \ 

v 

















































































































































































8 



/O // /Z '3 



/ z 3 y s (o 7 
ID 

Fig. 2. Graph showing deviation from the mean of each of the components 
in a mixture of equal parts. N, number of deviates in 50 tests; D, percentage 
of deviation in 100 seeds. 



paper in the standard germinating chamber at alternating tempera- 
ture, and each set of 50 was made at the same time by the same per- 
son. For sake of comparison, the results of four lots tested by 
about 20 different laboratories (Stone, 191 3), have been calculated 
and added. These figures are shown in Table 3. 



Table 3. — Probable error in gcrmination'tests. 



Kind of seed. 


No. of tests. 


Percentage of germination. 


Probable error in test 
of 100 seeds. 




50 


50 (48 hard) 


4-03 


Alfalfa (2d trial) 


50 


50 (48 hard) 


3-35 


Alfalfa No. 2 


50 


92 (2 hard) 


1.89 


Alfalfa No. 3 


50 


85 (i3.5 hard) 


1. 81 


Alfalfa No. a 


50 


91 (4 hard) 


2.14 


Millet 


50 


9i 


2.58 




50 


97-5 


1. 18 


Alfalfa 


19 


96 


. 1,99 




18 


9i 


2-43 


Millet ' 


18 


93 


1.97 


Timothy 


18 


92 


2.20 



8 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



The preceding work is based upon only 100 seeds for each test. 
The next question is to what extent the error would be reduced by 
using a greater number. For this purpose several series of 200 tests 
each were made in the same manner as before, but each lot of 50 
was run at a different time. The results have been calculated for 
each set of 50 tests, then for each set of 50 obtained by taking the 
mean of the first and second, third and fourth, etc., and in the same 
manner for the means of each set of three and four tests, thus giving 
series based upon 200, 300, and 400 seeds respectively. The results 
of these are shown in Table 4. 



Table 4. — Probable error in using 100, 200, 300 and 400 seeds. 



Kind of seed. 


Percent of 
germination. 


Number of seeds in test. 


100. 


200. 


300. 


400. 


1st set. 


2d set. 


3d set. 


4thset. 


1st set. 


2d set. 


Millet 


72.O 


3-51 


2.98 


3-09 


a 


2-47 


a 


2.25 




Alfalfa 


83.0 (9 hard) 


2.49 


2.62 


2.48 


3.02 


3-01 


2.27 


1.82 


1-43 


Do 


56.S 

(35-5 hard) 


4.10 


3.63 


3-32 


3-51 


2.79 


2-53 


2.72 


1 .91 


Flax 


99.0 


.68 


.81 


.61 


•79 


•47 


•52 


•37 


•37 


Kafir 


b 6o.o (32 hard) 


3-53 


3-78 


3-24 


3.26 


2.72 


2.26 


2.10 


1.82 



Only 150 tests made. 

6 Not germinated ; mixture as in Table 2. 



Vs this method of comparing such data may not be considered per- 
missible, an 80 percent mixture was prepared and a further series of 
5rxj lots of 100 each were counted. The rata thus secured were cal- 
culated in four distinct groups and also in combinations as in the 
former cases, with the following results. 

For 100 seeds — 2.65 (1st lot of 50) ; 

For 200 seeds — 1.78 (2d and 3d lots used as 50 of 200 seeds) ; 
l-r.r 300 "-reds 1.50 Cjth to 6th lots used as 50 of 300 seeds); 
For 400 seeds — 1.35 (7th to 10th lots used as 50 of 400 seeds). 
For each of 10 lots of 50 of 100 seeds each — 2.65, 2.55, 3.1 1, 2.70, 3.21, 

2.36, 2.05, 2.41, 2.14, 2.96. 
For each of 5 lots of 50 of 200 seeds each — 1.69, 1.76, 2.13, 1.77, 1.88. 
Pot each "f 3 lots of $0 of 300 seeds each — 1.64, 1.50, 1.34. 
Poff etdl of 2 lots of JO '»f 400 seeds each — 1. 10, 1.35. 
For each of 2 lots of 50 of 500 seeds each — 1.08, 1.25. 
For 1 lot of i t ooo seeds each — 0.85. 

Figtlft 3 ihowi ili<- deviation of joo ]<.|s in the third and fifth series 
of Table 4 and aNo for the 500 last mentioned. This figure shows 
how little the variation at 50 percenl differs from that at 80 

percent. 



STEVENS : VARIATIONS IN SEED TESTS. 



9 



By using less than 100 seeds in each lot the error would be con- 
siderably increased. One trial of a 50 percent mixture, using 50 lots 
of 50 and 25 seeds each, gave probable errors of 4.26 and 6.73 re- 
spectively. 

Another question arises in this connection. If the results of two 
tests vary rather widely, does it follow that the mean of the two is 
farther from the mean of the series than when the variation between 




I 2 



23 



7 /o 

JD 

Fig. 3. Deviation from the mean of 200 and 500 lots of 100 seeds each. P , 
percentage of deviates ; D, percentage of deviation ; a, alfalfa having 56.5 per- 
cent viable and 35.5 percent hard seeds ; b, Kafir mixture, 60 percent calculated 
viable and 32 percent hard; c, Kafir mixture, 80 percent calculated viable and 
20 percent dead. 

two tests is less? In other words, is it necessary to make a retest 
when more than a certain variation occurs between duplicate tests? 

To throw some light upon this point the second and fifth series in 
Table 4 were used. The mean of each successive pair (1st and 2d, 
3d and 4th, etc.) was found and the deviation of these means from 
the mean of the entire series of 200 was found. When these dif- 
ferences are arranged according to the variation of the two tests of 
each pair the results shown in Table 5 are produced. 

The data from the last series of 500 tests were also handled in the 
same way and in addition the percentage of cases where the mean of 
the pair exceeded the mean of the entire series by more than the 
probable error of 200 seeds was found for each group. These data 
are given in Table 6. 



10 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Table 5. — Deviation of duplicate tests. 



Variation between 
the members of 
each pair. 


Deviation of mean of each pair from mean of entire series. 


Alfalfa 


series. 


Kafir series. 


Number of deviates 
* um er eviates 


Avera e deviation 
\erage e\iation. 


Number of deviates. 


Average deviation. 






Percent. 




Percent. 





2 


2.0 


s 


1.6 


I 


7 


1 7 


9 


2.8 


2 


I J 


2 8 


9 


2.7 


3 


O 

y 


2.8 


13 


2-3 


4 



y 


2 7 


10 


3-3 


5 


1 


2.7 


9 


3-6 


6 


T H 

1 / 


3-5 


14 


2.4 


7 


g 


2.9 


6 


4.8 


8 


2 


i n 
J • {J 


8 


2.1 


9 


6 


4.4 


4 


5-4 


10 


2 


6.0 


4 


4-5 


11 


6 


4-5 


1 


1-3 


12 


1 


2-5 


1 


1.2 


13 


2 


3-5 


1 


3.3 


14 


5 


4.2 


3 


3 4 


17 


1 


4.0 


1 


3-3 


18 







1 


.8 


21 







1 


1-7 


22 


1 


i-5 







24 


1 


7-5 








Table 6,— Variation of duplicates in 500 lots of 100 seeds each, counted from 

an 80 percent mixture. 



Difference 


Number of 


Average 


Number 


Percentage exceeding probable error. 


between 


exceeding 








duplicates. 


cases. 


deviation. 


probable error. 


Individuals. 


3 groups. 


2 groups. 


Percent. 




Percent. 













17 


2.0 


I I 


05 












40 


2.0 


12 


30 












29 


1.8 


14 


48 




45 








27 


2.1 


IO 


37 








48 


4 


25 


2.2 


IS 


60 












22 


2.4 


II 


50 












23 


2 5 


is 


67 












13 


1-7 


2 


IS 




52 








IO 


1-5 


5 


50 












7 


3-7 





B6 










10 


1 2 


2.0 




33 








■ 46 


1 1 


9 


3 7 


i 


<>1 




49 






12 


9 


2.2 


4 


44 










.3 


* 


94 




50 










14 


3 


1-7 


9 


'<7 











Thm is at present a ruling adopter] l> v the Official Seed Analysts 
of North America that retests should he made when the variation 
in duplicate tests of 100 seeds each exceeds the following: 



STEVENS : VARIATIONS IN SEED TESTS. 



6 percent for a germination of 90 percent or more; 

7 percent for a germination of 80 to 90 percent; 

8 percent for a germination of 70 to 80 percent; 

9 percent for a germination of 60 to 70 percent; 
10 percent for a germination of 50 to 60 percent. 

The writer believes that this regulation is of doubtful value in that 
it is quite possible for the variation to be greater than this without 
destroying the value of the test ; further, that it is likely to cause an 
unwarranted faith in results which show smaller variations. 

From Table 6 it will be observed that slightly over 20 percent of 
the series would be subject to retest by the above ruling. The fifth 
column of the table shows the percentage of cases in which the mean 
of each pair of tests exceeds the mean of the entire series by more 
than the probable error o'f a test of 200 seeds. Since by the nature of 
the probable error it is exceeded by half of the cases, the amounts 
in this column should average about 50 percent. If the duplicates 
which differ more than 7 percent are not to be relied upon, we should 
expect to see the values in this column higher for variations above 
7 percent than for those below. It will be seen from the table that 
this is not the case ; that approximately the same number of high and 
low values occur in different portions of the column, and that by 
collecting the results in two or three groups, very little difference is 
found. 

Of the 250 duplicates, 46 showed a difference of over 8 percent 
between the two members. The average deviation o'f the means of 
these 46 from the mean of the entire series was 2.35 percent, while 
in the other 204 cases it was 2.06 percent. Further, in only 24 of the 
46 was this deviation greater than 2 percent (probable error in test 
of 200 seeds is 1.87 percent). This seems to show plainly that in this 
case wide variations between duplicate lots did not appreciably reduce 
the accuracy of the result. 

In Table 5, 9 of the 100 duplicates of the kafir series showed a dif- 
ference o'f over 10 percent between the mean of the two tests and the 
mean of the entire series. The average deviation of the means of 
these 9 from the mean of the entire series was 2.44 percent; that of 
the other 91, 2.97 percent. Only in 4 of the 9 did this deviation ex- 
ceed the probable error of 2.2 percent. 

In the alfalfa series given in the same table, 17 of 100 exceeded the 
limit of 10 percent, the average deviation being 4.15 percent, and 12 
of the 17 exceeding the probable error. The average deviation of 
the other 83 was 3.44 percent. 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



The results of this alfalfa series seem to" oppose the writer's con- 
tention, yet the data are few and the kafir series tends the other way 
to about an equal extent. It is to be noted further than among- dupli- 
cates showing- only slight variations the means may differ widely from 
the mean of the series. For instance, in the alfalfa series just men- 
tioned there is one case among- the duplicates which vary only I 
percent in which the mean of the two differs from the mean of the 
series by 9 percent ; among those differing by 2 percent, there are 
two showing 6.5 percent and one 5.5 percent, etc. 

This is a point which the writer wishes to' emphasize especially. 
If duplicates vary only slightly, it is still uncertain that their mean is 
accurate to a similar degree. If the conditions for the test are un- 
suitable this will be even more the case, for if one test fails to 
germinate properly the chances are that the duplicate will do the same. 

Variations in Purity Tests. 

A large number of factors are concerned in the variations in 
sampling for purity tests, since each of the three chief components 
of the sample (foreign seed, inert matter, and pure seed) may have 
a variable number of components. The size and number of the 
various foreign seeds will play an important part. Thus it may be 
difficult to find a value of accuracy which shall serve for a large 
range of cases. 

In the following experiments the general plan was the same as in 
the preceding, viz., 50 tests of each lot made by the same person, 
the seed U9ed being such that unusual variations would not be en- 
COlintered. In taking the standard samples for analysis the seed 
mixer and sampler 4 was used, but the last separation was corrected 
to the standard weight by dipping out with a spoon or adding in the 
same manner from the other portion of the separation. The entire 
lot was not separated eaeh time, enough for several tests being set 
toward tlx- end of the process, this finished, and a similar amount 

■' aside from Another separation of the bulk. The bromegrass, 

although a fairly clean lot, would not run through the mixer, and was 

poured into a pan, portions being taken from different parts for each 

lot tet)t6d The resttltl are shown in Table 7. 

'I he large error in foreign seed of the bromegrass is due to occa- 
sional seeds of barley and wild oats. Such an occurrence is common 
in the homegrown bromegrass seed examined by us. Similar condi- 

• Rulr» ami api>.«o.tM, f.,i ■ < <\ 1. tm«. I '. S. Dept. A«r ., Office of Kxpcri- 
mcnt Stations Circ. 34, rev. cd., p. 12, KJ04, 



STEVENS : VARIATIONS IN SEED TESTS. I 3 



Table 7. — Probable error of purity tests as determined with the seeds of 

several crops. 



Kind of seed. 


Number 
of tests. 


Quantity 
used. 


Pure seed. 


Foreign seed. 


Inert matter. 






Grams. 


Percent. 


Percent. 


Percent. 


Flax No. 1 


50 


10 


97-88± .15 


0.63=*= .06 


1.49=*=. 13 


Flax No. 2 


SO 


10 


97.i8± .17 


1.17=*=. 13 


i.;65*.i3 


Alfalfa 


50 


5 


98.00=*= .17 


1. 56 ±.1.6 


.44=*=. 07 




So 


3 


89.40=1= 1. 00 


.28=1= .32 


io.i8±.87 



tions may be met elsewhere, and in ordinary work the result should 
be corrected from the examination of a larger quantity for such im- 
purities, if it is apparent that the result would be materially changed 
thereby. 

In all but the last of the above series the percentages were cal- 
culated to the second decimal. To show further the value of this, 
the probable error was calculated for the first two a second time, 
using only one decimal, with the results shown in Table 8. 



Table 8. — Value of second decimal in calculating probable error of purity tests. 







Probable error. 


Test. 


Percentage. 










Using two decimals. 


Using one decimal. 


Flax No. I. 




Percent. 


Percent. 




O.63 


0.062 


0.068 


Inert matter 


I.50 


.129 


.147 


Pure seed 


97-88 


•155 


•157 


Flax No. 2 










I.I7 


.132 


.119 


Inert matter 


I.65 


• 133 


•143 




97-18 


.171 


.172 



The variation in number of foreign seeds of a given species from 
the tests reported in Table 7 is shown in Table 9. 



Table q. — Variation in number of foreign 


seeds 


in purity tests. 


Seed examined and foreign seed contained therein. Average number. Probable error 


Flax No. 1 : 








29.3 


3-8 




9-4 


1.6 




1.2 


7 




• 7 


•5 


Flax No. 2: 








14.0 


2.1 




. 36.6 


4-4 


Alfalfa : 








88.5 


7-4 




1.2 


.8 




• 17 


1.0 



14 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Seed examined and foreign seed contained therein. Average number. Probable error. 

Bromegrass : 

Agropyron tencrum 1.4 .8 

Lappula Iappula 3 -4 

Polygonum convolvulus 4 .4 



One attempt was made to ascertain the variation due to persona! 
selection. The last lot in the first flax series was thrown back 
together and reworked three times by three analysts, the weighing of 
each separation being made by one of the other analysts in order to 
avoid unconscious selection in the next test. The results are pre- 
sented in Table 10. 



Table 10. — Variation in separation of the same sample by different analysts. 



Analyst. 


Foreign seed 




Inert matter. 
















Test No. 1. 


Test No. 2. 


Test No. 3. 


Test No. 1. 


Test No. 2. 


Test No. 3. 


No. I 


•71 


.70 


•71 


1-34 


1. 12 


1-33 




.72 


•71 


•73 


1.25 


I.20 


1.47 


No. 3 


.72 


.72 


.72 


1. 14 


1.20 


I.46 



A.S the foreign seeds were identical in each case the difference in 
those columns must be due to inaccuracy in weighing. The balance 
used was an inexpensive one, but the variation is not of significant 
value. In case of the inert matter one factor is the distinction be- 
tween various conditions of broken seeds, these comprising most of 
\}]<- quantity. Further tests of this sort are desirable, but they are 
difficult to control properly. 

Conclusions and Recommendations. 
1. The probable error of a single germination test of 100 to 400 
seeds vario according to percentage of germination as shown in 
Table i 1 . 

TaBLI 11. — Approximate probable error for germination tests. 



Vert muffa of pnmtaatioii. 



ami. 


2 


5 


95. 


yo. 


80 to 50. 


100 


•7S 


1 .Of) 


1.50 


2.25 


2. SO 


200 




■ )" 




" I.JO 


2.0U 




.40 


"55 


" So 


" 1 .20 


1-75 






* - So 


.70 


" 1. 05 


I.50 



m.itrfl from \\u value next aliovc l>y reducing il in proportion to the 
in the first column. 



STEVENS : VARIATIONS IN SEED TESTS. 



15 



The above figures are increased about one-fifth in the lower per- 
centages of germination for legumes containing the so-called "hard" 
seeds. This is for work in which the sources of variation are reduced 
as far as it is possible to do. No attempt is made in this paper to 
determine the range of variation where factors other than that of 
mathematical probability enter to any extent. These values may also 
be used for other experiments involving similar conditions, e. g., 
counting 500 seeds to determine percentage of mixture of two kinds. 

2. For samples not containing mixtures of materials which tend 
to separate readily (such as sand, fine trash, or coarse material), 
only a small amount of mixing of samples seems necessary. Samples 
which do contain such mixtures should receive, when practical, a 
supplementary test of larger quantity to show the approximate quan- 
tity of such materials. For example, these may be separated first by 
a sieve, and the percentage added to that obtained by a regular test 
from the remaining quantity. 

3. For purity tests the accuracy depends upon many factors. The 
quantities used should receive a careful investigation in order to 
determine whether those in current use could be changed to advan- 
tage. The use of the second decimal place is of no value in most 
work. If such accuracy is desired, the test should be based upon a 
sample of sufficient size. 

An instance of such change may be cited in some work carried on 
at this laboratory in connection with the Improved Seed-Growers' 
contest. The following schedule was adopted and the second decimal 
used. For cereals a measured quantity (about 8 ounces for wheat) 
was used unless the seed was obviously impure, and then the regular 
quantity (30 gr.) was taken. For flax and smaller seeds, three times 
the regular quantity was taken unless decidedly impure. This is not 
suggested as a new basis, but merely mentioned as an instance of a 
sliding scale which has been used to advantage. In this work the 
number of seeds of certain noxious weeds, such as wild oats, mustard, 
etc., was calculated from the larger sample. This is quite an im- 
portant point and should be carried out for any sorts that are con- 
sidered of special importance. 

4. Results of seed tests should be accompanied by an indication of 
their accuracy. The regular way of expressing this is by writing 
the probable error after the result, e. g., 95 ± 1.5 percent, thus in- 
dicating that the result probably lies between 93.5 percent and 96.5 
percent. For ordinary reports it would be desirable to have some 
form by which it would be stated more completely. 



1 6 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



While the probable error represents a very definite quantity, so 
far as the data from which it is derived are concerned, its practical 
application is somewhat difficult. Using again the figures of the 
preceding paragraph, the probable error represents only an even 
chance that the true result lies within 1.5 percent of 95 percent. 
The chances are as great that it lies beyond this. If the probable 
error is doubled, there are about four chances in five that it lies 
within this figure (95 ± 3, i. e., between 92 and 98). Placing this 
in another form we see that by allowing twice the probable error, our 
results will still be beyond this in 20 percent of the trials. This is 
surely the smallest reasonable allowance that can be made. Then for 
the smallest practical scale of allowance for variation, the figures in 
Table 1 1 should be doubled. 

5. Table 8 indicates that the second decimal place is not necessary 
for the calculation of probable error in such tests. This is quite im- 
portant as the use of the second place involves several times as much 
labor in calculating. 

6. The amount of seed used for tests (and therefore the degree of 
accuracy obtained) must be regulated by two factors, viz., the degree 
of accuracy necessary for dependable results and the amount of work 
which it is possible to handle. From the data presented in Tables 
5 and 6 it would seem that for germination tests, 200 seeds in a single 
test would be advisable for ordinary work, the number being in- 
creased when desired. It is very important that the probable error 
be known so that such adjustments may be made. 

7. Duplicate tests appear to be of little value as, so long as only 
the factor of probability in selection is present, variation between 
duplicates is not significant; if other factors enter, the chances are 
probably as ^reat that duplicates which vary but little are unreliable. 
The necessity of making another test must be governed chiefly by 
judgment, whether duplicates vary or not, and a test of 200 seeds 
will often require less time and space than two of 100 each. 

( QMPAB150N WITH TBB RESULTS OF RoDEWALD. 
the work of I )r. II. Uodcwald came to notice after the present 
papei ,va practically completed, il ■>ccms desirable to add some cx- 
ihowing his results. An abstract of the earlier paper which 
i- a practical^ complete translation has been published ("(i. M. €.," 
iH/yi . Rodewald states that errors are either accidental or system- 
\ imilar statement ua- made in the beginning of the present 
paper and in very nearly all of the writer's work an attempt was made 



STEVENS: VARIATIONS IN SEED TESTS. 



I/- 



to reduce the systematic errors to a minimum or constant quantity, in 
order to determine the accidental error. 

The data in Table 12, given by Rodewald, were determined by 
mathematical calculation. 



Table 12.— Probable error due to accident. (Rodewald, 1889, p. no.) 



Percentage of 
germination. 



Percentage of probable error in using 



100 seeds. 



95 
go 
80 
70 
60 
So 



1.48 
2.03 
2.71 
3-H 
3 32 
3-39 



200 seeds. 



300 seeds. 400 seeds. 



I.04 
I.44 
I.92 
2.20 

2-34 
2.4O 



O.85 
1. 17 
1.56 
1.79 
I.92 
I.96 



O.74 
1.02 
1.36 

i-55 
1.66 
1.69 



500 seeds. 



900 seeds. 



0.66 
.91 
1. 21 

i-39 
1.49 
1.52 



0.49 
.68 
.90 
1.04 
1.1 1 
I-I3 



Comparing these results with column I of Table 2 and with Table 
11, they are found to be almost identical, except that the former 
shows approximately equal values for 80 percent to 50 percent when 
only two factors are considered. Where the third factor is con- 
sidered the results conform closely to those given above (indicated 
in figure 1). Rodewald apparently did not distinguish between these 
two conditions, nor has he given values for percentages lying between 
95 and 100. He found these theoretical values to conform closely 
with the results from actual tests, but for this he used the results of 
a number of samples showing approximately the same germination 
instead of making a large number of tests from one sample. 

From several experiments in which results of different labora- 
tories were compared, he found that the systematic error held a 
rather constant relation to the accidental. From this he finds that 
the total error is about 2.2 times the accidental error (Rodewald, 
1904). From this he derives the following table of variation which 
should not be exceeded in more than 4.3 percent of comparative tests 
of 200 seeds each by different laboratories : 



For 95% 
For 90% 
For 85% 
For 80% 



germination, 
germination, 
germination, 
germination, 



For 75% germination, 



6.3% ; For 70% germination, 14.4% ; 

9.4%; For 65% germination, 15.0%; 

11.2%; For 60% germination, 154%; 

12.6%; For 55% germination, 15.7%; 

13.6%; For 50% germination, I57%- 



For allowable variation between duplicate tests he gives the fol- 
lowing, which will still be exceeded in 4.3 percent of the cases (Rode- 
wald, 1889, P- 22 6) : 



i8 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Percentage of 
germination. 



Allowable variation in each test. 



Using 200 seeds, 



Using 100 seeds. 



95 
90 
80 
70 
60 
50 



4.62 

6.49 
7.62 
9.78 
10.38 
10.65 



6-57 
9.04 
1306 
13.82. 
1477 
15.08 



The variation allowed here between two tests at 80 percent germi- 
nation is but slightly less than the extreme variation between dupli- 
cates shown in Table 6. The extent to which the results of Table 5 
exceed such allowance is slightly greater but still within the 4.3 
percent limit of accuracy of the table. This is surely a confirmation 
of the writer's contention that duplicates varying 10 percent, 15 
percent, or even more may still be reliable tests. 

For purity tests Rodewald considered that an allowance of 2 per- 
cent could be used for clover samples having a purity of 90 percent 
or more, and 3 percent for those below 90 percent, and that, similar 
values might serve for grasses, if the purity is determined by the 
weight method (Rodewald, 1904, p. 113). The mean and probable 
error of tests by 23 stations was 97.992 ± .406 percent for one sample 
of red clover, and 70.032 ± 1.288 percent for one of orchard grass. 
The present writer believes that the space between 90 percent and 100 
I »erceilt should be further divided, and it should also be borne in mind 
thai the amounts used for purity determination in Rodewald's tests 
wen- twice as large as those in current use in the United States. 

The writer has quoted thus extensively from this work partly for 
sake of comparison and partly to give some of the results, as the 
paper* themselves may be relatively inaccessible to many other 
analysts, a- they were to me. My thanks are due to Mr. J. P. Helyar 
of the New Jersey station and to Mr. Edgar Brown of the U. S. 
Department Of Agriculture for calling my attention to them. Por- 
tionj of the later paper referring to attempts to reduce the systematic 
ermr and comparison of weight and count methods have not been 
ed, but it may be well to quote a few statements from the 
summary : 

P01 tin clover sorts both weight and court methods are about 0$ equal value. 

I - r tine grasses both have large systematic errors. It would seem that the 
\\«:^'ht method would |>c the more easily carried out (p. 113). 

1 he fihyMolfigiral ran t s for the large systematic errors of germination tcst- 
inK are yet unexplained. It appears that the evaporating conditions play an 
important role (p. 117). 

Ih« laryi rtroi <,\ <t<\\> \ \ Ui . are not to he sought in the less worthy work 
Of the M-#-d rontrol Mations. 'I In \ are grounded in the nature of the case. 



dunnewald: fertility of sandy pine plain soils. 19 
Literature Cited. 

" G. M. C." 

1891. The theory of probability applied to seed testing. In Agr. Science, 
5 : 74-77, 96-105. 

RODEWALD, H. 

1899. tiber die Fehler der Keimpriifungen. In Landw. Vers. Stat., 36: 
105-112, 215-227. 

1904. Untersuchungen iiber die Fehler der Samenpriifungen. In Arb. 
Deut. Landw. Gesselschaft, Heft ioi, p. iv + 117. 

Stone, G. E. 

1913. Some variable results of seed testing. In Ann. Rept. Mass. Agr. 
Expt. Sta. 1912, part 2, p. 22-30. 



VEGETATION AS AN INDICATOR OF THE FERTILITY OF 
SANDY PINE PLAIN SOILS IN NORTHERN WISCONSIN. 1 

T. J. Dunnewald. 

While making a survey and report on the soils of a proposed Forest 
Reserve area in northern Wisconsin, it was noticed that the sandy 
plains soils varied greatly in their ability to produce a second growth 
of vegetation after the removal of the original pine timber and the 
many severe fires which succeeded the logging operations. 

The most sandy portions where the original timber was sparse or 
consisted mostly of Norway and Jack pine, with perhaps a few white 
pines, now bear little or no second growth. Small Jack or Norway 
pines 6 to 10 feet high appear in clumps and the poplar brush, if any, 
is also less than 10 feet high, while a thick growth of sweetfern, 
brakes, blueberries, or coarse bunch grass is the only ground cover, 
In other places where moisture conditions appeared somewhat better 
and the soil slightly more loamy, the second growth is often 20 to 
40 feet high and consists of poplars, white birch, cherry, alder, and 
young white pine, with but few Jack or Norway pines. The original 
timber also had been of a better quality here, being mostly large white 
and Norway pine, as indicated by the stumps. 

In the final correlation of the soils on the basis of their value for 
agricultural purposes, the most sandy soil was described as being 
of low value for farming, while the more loamy soil, as indicated by 
the vegetation and better moisture conditions, was classed as being 

1 Contribution from the Wisconsin State Soil Survey, Madison, Wis. Re- 
ceived for publication April 6, 1917. 



20 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



of fair value for future farming. Typical samples from widely. 

separated areas of these classes of soil were collected and analyses 
made to correct or confirm the field interpretation. 

Table I gives a summary of the various determinations made on 
these two groups of soil samples. 

Table i. — Chemical and mechanical analyses of sandy pine plain soils bearing 
good and poor second growth. 

SOILS WITH SPARSE SECOND GROWTH. 



Chemical analyses. 



Mechanical analyses. 



SoilNo. 


P. 


K. 


N. 


Lime Moist- 
re- ure 
quire- equiv- 
alent." alent. 


Gravel. 


Coarse 
sand. 


Me- 
dium 
sand. 


Fine 
sand. 


Very 
fine 
sand. 


Silt. 


Clay. 


Class of 
soil. 














Pel. 


Pet. 


Pet. 


Pel. 


Pet. 


Pet. 


Pet. 




919 .4 


0.052 


I.24 


0.096 


2,000 


15-3 


6.85 


22.44 


22.20 


27.08 


46.90 


9.61 


6.92 


Coarse 


919 B 


•035 


1.36 


•045 


500 


7-4 


9.10 


18.43 


21.58 


37-37 


5-42 


4-74 


3-35 


Medium 


882 A 


.052 


1.29 


.063 


2,000 


14.8 
















do. 


882 B 


.041 


1. 12 


•033 


2,000 


9.8 
















do. 


873 -4 


.051 


O.76 


•095 


2,000 


15.0 
















do. 


873 B 


.031 


0-95 


.027 


500 


9-3 
















do. 


861 A 


.042 


I.I I 


•057 


2,000 


14.9 
















do. 


861 B 


.027 


.87 


.056 


2,000 


ti-S 
















do. 


859 A 










16.7 
















Medium 
to fine 


859 B 










12.0 
















Fine 


743 A 


.042 


1. 19 


.047 


2,000 


12.8 


2-54 


15-79 


29-45 


28.54 


6.84 


8.49 


5-76 


Medium 


743 B 


.029 


I .OI 


.025 


500 


6.6 


4-03 


14-57 


30.30 


31.11 


9.87 


6.17 


4-13 


do. 



SOILS WITH LARGE SECOND GROWTH. 



863 A 


.059 1.06 


.099 


2,000 


I9.9 


1.25 


4-33 


8.40 


46.17 


21.18 


12.371 6.32 


Fine 


863 h 


•033 


E.Xfl 


.049 


2,000 


13.8 


2.00 


4-13 


9.20 


51.01 


21.62 


7-95 


4.18 


do. 


862 A 


.041 


0.97 


.079 


2,000 


16.5 
















do. 


862 B 


.032 


O.97 


.032 


2,000 


13.5 
















do. 


M<; A 


.047 


1. 21 


.082 


2,000 


21.9 
















do. 


* V) B 


.037 1. 17 


.049 


2,000 


l<>.~ 
















do. 


7K0 4 


.058 1. 12 


.078 


2,000 


22.9 


7-23 


20.05 


20.82 


14.49 


8.22 


20.98 


8.80 


Medium 


780 B 


•039 


I.3I 


037 


J. 000 


15.2 


















785 A 


.068 


1. 21 


.094 


5,000 


18.4 


0.85 


1 i .07 


[9.O4 


2 7-93 


27.06 


1 1.26 


3.20 


Medium 
to fine 


785 B 


.042 


MS 


.028 


5.000 


12.7 


0.80 


0-44 


18.33 


28.51 


32.08 


7-47 


2.64 


Medium 
to fine 


741 A 


.063 


O.87 


.068 
.026 


2.000 
2,000 


\(>.() 
10.6 
















Medium 
to fine 

Pine 



k Ltn( requirement in potindl <»f calcium carbonate necessary to neutralize 
t!.< ur fa< '• X inches of soil. 



A mechanical separation of the particles of different sizes of which 
tflC toil are composed was made OXl several of the samples. The 
r<- nil . a will he seen in Table 1, indicate that the soils with small 
or spai e (Mid growth must he classed as coarse or medium sand, 



dunnewald: fertility of sandy pine plain SOILS. 2 1 



while those bearing large second growth should be classed as fine 
sand because they contain larger amounts of the smaller sized soil 
particles and fewer of those of larger size. The curves in figure 4 
show the total average percentages of soil particles of different sizes 
in the two groups of samples as shown by the mechanical analyses. 

Tfcf 









-v-V 








^ 


1 

*s 1 


\ \ 
\ \ 








s 

/ 

/ 




— \ — V 
\ \ 
\ 






/ 

/ 


/ 




V 

\ 






/ 

— M 






\ — 

\ 
\ 












v— 

\ 

\ 



















FINE GRMEL COARSE SAND WEDIUN SAND FINE SAND SILT CLAY 

Fig. 4. Graph showing total average percentages of soil particles of different 
sizes in the two groups of samples (Table 1). 



Another arrangement of the percentage in Table 2 indicates the 
predominance of the finer particles in the soils bearing the best 
vegetation. It will be seen at a glance that while the content of the 
finest clay particles is much the same, the silt and especially the 
finer sands are present in much greater proportion in the soil where 
the better reproduction is found. 



Table 2.— Percentage of silt, clay, and finest sand particles in pine plain soils 
of northern Wisconsin. 



Sandy soil, 
stunted vege- 
tation. 


Clay. 


Silt^ 


Fine and 
very fine 
sand. 




Percent. 


Percent. 


Percent. 


919-4 


6.92 


9.61 


31-77 


919 B 


3-35 


4-74 


42.79 


743 -4 


5-76 


8.49 


35-38 


743 B 


4-13 


6.17 


40.98 


Averages .... 


504 


7-25 


37-73 



Finer soil, large 
vegetation. 



863 B 
785 A 
785 B 







Fine and 


Clay. 


Silt. 


very fine 




sand. 


Percent. 


Percent. 


Percent. 


6.32 


12.37 


67-35 


4.18 


7-95 


72.63 


3.20 


11.26 


54-99 


2.64 


7-47 


61.49 


4.08 


9.76 


64.11 



2 2 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Chemical analyses for total phosphorus, nitrogen, potassium, 
organic matter, and calcium were made on the samples collected. 
The individual determinations given in Table I are averaged for the 
groups in Table 3. 



Table 3. — Percentage of phosphorus, nitrogen, potassium and phosphorus in 
loamy and fine sands as compared with the coarser sands. 



Portion ol soil. 


Coarse sands. 


Loamy and fine sands. 


Phosphorus. 


Potassium. 


Nitrogen. 


Phosphorus. 


Potassium. 


Nitrogen. 




Percent. 


Percent. 


Percent. 


Percent. 


Percent. 


Percent. 


Surface 


0.048 


1. 12 


0.071 


0.055 


1.05 


0.081 


Subsoil 


0.032 


1.06 


0.037 


0.037 


1. 13 


0.035 



The amounts of the plant food elements, phosphorus, potassium, 
and nitrogen, are low in all these samples as compared with other 
classes of soil which contain more of the finer soil material, such as 
clay and silt. The greatest difference between the groups is in the 
element phosphorus, of which the sandy group have 14 percent less 
in the surface 8 inches of soil. The element nitrogen is present also 
in about 14 percent smaller amount in the surface soil of the sandy 
group than in the loamy soils. Other chemical data not given in the 
table show that the total calcium is but 0.82 percent in the sandy 
group and 1.16 percent in the loamy group. Expressed as pounds per 
acre of soil 8 inches deep the food elements are as shown in Table 4. 



TABLE 4. — Pounds per acre of phosphorus, potassium, and nitrogen in certain 
Wisconsin soils to a depth of 8 inches. 



Portion of soil. 


Low agricultural value. 


Fair agricultural value. 


Phosphorus. 


Potassium. 


Nitrogen. 


Phosphorus. 


Potassium. 


Nitrogen. 




895 
C).J2 


20,2Ko 
21.653 


I.39I 
685 


1,170 
855 


26,400 
25.530 


3.147 
1,151 



The moisture equivalent determination consists briefly in placing 
filial bulks of the different soils in perforated cups in a centrifuge 
machine, after they have absorbed all the water they will hold. They 
ar<- then lllbjected to a Speed of 2,440 revolutions per minute for 40 
" ' ' '' 'I'd 'f«- percentage of moisture remaining after treatment is 
called the moisture equivalent. The determination is intended to 
a Comparative figure lot the moisture-holding capacity of dif- 
■oil 'I lie averages of the determinations are shown in 

Tabic 5. 



freeman: hard and soft kernels. IX WHEAT. 



25 



Table 5.— Average moisture coefficients of sands and of loamy sand soils in 

northern Wisconsin. 

Fine and loamy 

Portion of soil. Sands. sands. 

Surface soil 1492 19.40 

Subsoil 9.48 13.76 

The group of finer samples, as will be seen, have 27 percent (nearly 
one-third) greater capacity for holding moisture than have the sands. 
The difference is about equally divided between the surface and sub- 
soil. The greater capacity of the surface 8 inches of both groups 
to hold moisture as compared with their respective subsoils may be 
attributed to the organic matter accumulated in the surface 2 to 3 
inches of soil. This material, derived from the partial decay of 
vegetation, has a very high water-holding capacity. This greater 
water-holding capacity is an especially important matter in judging 
of the crop value of sandy soil and the lack of the larger water- 
holding capacity often means the loss of crops on the more sandy 
soils during even short periods of drought. 

Conclusions. 

It is concluded that the character and size of the undergrowth of 
cut-over lands is a safe indicator of the cropping capacity of the 
soil for agricultural purposes on sandy pine plain lands. 

The heavier growth indicates a higher content of plant food, the 
presence of more fine material in the soil, and especially a greater 
capacity of the soil to retain moisture and to enable vegetation and 
future crops to resist periods of drought. 



A MECHANICAL EXPLANATION OF PROGRESSIVE CHANGES 
IN THE PROPORTIONS OF HARD AND SOFT KERNELS 

IN WHEAT. 1 

Geo. F. Freeman. 

Soft wheats of low gluten content are usually found in warm humid 
or irrigated sections. Many of these varieties when taken to drier, 
colder climates, produce hard, translucent grains having the horny 

1 Contribution from the University of Arizona, Tucson. Ariz. Read for the 
writer by Dr. R. H. Forbes at the Second Annual Conference of Agronomic 
Workers in the Eleven Western States at Pullman, Wash.. August 2, 1917. 
Received for publication August 27, 1917- 



24 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



texture of wheats characteristic of these conditions. Other varieties 
produce soft, opaque grains in all situations and there are yet other 
sorts which produce a greater or less percentage of hard, translucent 
grains even in warm, humid regions. 

With reference to their response to environic conditions, Howard, 
Leake, and Howard- separate wheat varieties into three classes, as 
follows: (i) Wheats which always remain soft, (2) wheats with 
a tendency to remain hard, and (3) the majority of varieties in which 
the consistency varies greatly according to the locality and the con- 
ditions under which they are grown. An investigator working upon 
the relative effects of variety and of climate on the quality of wheat 
might easily be led to serious errors of conclusion should he chance 
to include in his tests varieties belonging to only one of these classes. 
Particularly would this error impend should he work only with class 
3. It is not to be understood that the three classes here defined have 
well marked limits. As a matter of fact they grade insensibly into 
each other through intermediate types. The three classes, therefore, 
merely distinguish between groups of hereditary tendencies which 
are more or less marked in their intensity. Hence, we have a field 
for almost infinite variety in the nature and sensitivity of the re- 
sponse to environment as expressed in the texture and composition 
of the wheat grain. 

\-. however, there are varieties which produce a large percentage 
of hard grains of high protein content even in warm, irrigated sec- 
tions and as these are qualities constantly sought by the millers, the 
question arises as to why the wheats of warm humid or irrigated sec- 
tions are usually of inferior grade. We have no conclusive evidence 
that there is any noticeable progressive change in the nature of the 
climatic reactions of a pure race of wheat when brought to a new 
environment. In view of the fact that hard wheats of high milling 
quality are often introduced into these localities with a view to im- 
proving the quality of the grain produced, why is it that these better 
varieties BO often are either lost or else soon deteriorate to such an 
extent th.'it they ;ire no better than the ordinary local sorts? The 
experiments now to be described had for their object the working 
out of the mechanism by which these better introductions are lost or 
else gradually deteriorate. 

During several year- in which the writer has had under observa- 
tion the wheats grown on the Arizona Agricultural Kxpcriment Sta- 
tion farm at Yuma, be lias noticed that commercial varieties of durum 

Howard \. Lcakr, II M. awl Howard, <>. < . Memoirs Dept. Ae,r. India, 
v. 5, no. 2. 1913. 



FREEMAN : HARD AND SOFT KERNELS IN WHEAT. 



25 



wheat have shown little or no tendency to gradually become softer 
from season to season but that both local and introduced commercial 
varieties of bread wheats are either entirely soft or else show the 
usual progressive tendency toward becoming soft. 

In the spring of 191 2 a large number of heads were selected and 
in the following fall all were sown in head rows. The pure races 
thus originated represented five types, as follows : White durum, 
red durum, poulard, red bread, and white bread wheats. All of these 
wheats were sown within one week's time, but their ripening periods 
spread out over about a month's time, extending from the middle of 
May to the middle of June. Hardness was determined by the per- 
centage of clear, translucent grains. No grain was considerd hard if 
it had even a spot of starchy, opaque endosperm. The different head 
rows showed practically every percentage of hardness from o to 100. 
Table 1 gives the average yield in grams per 100 feet of row and the 



Table i. — Correlation of yield and percentage of hard grains in various classes 
of wheat at Yuma, Ariz., in 1914. 



Class. 


Number of 
pure races. 


Average yield 
per 100 feet of 
row. 


Average per- 
centage of hard 
kernels. 


Correlation between 
yield and percentage of 
hard grains. 






Grams. 




Percent. 


White durum 


264 


2,032 


77 


+ 22 ± 4 


Red durum 


37 


1,481 


74 


+ 26 ± 9 


Poulard 


79 


1,636 


28 


— 26 ± 7 


Red bread 


179 


1.505 


53 


- 38 ± 4 


White bread 


65 


3. 131 


23 


- 52 ± 6 



percentage of hard grains in the five principal groups, together with 
the correlation between the yield and the percentage of hard grains 
in each group. In Table 2 the different classes are divided into hard 
and soft groups according to the percentage of hard kernels each 



Table 2. — Number of head rows in each class placed in the hard and the soft 
groups, with range in percentages, average percentages of hard grains, 
and yield to the 100 feet of row. 



Class. 


Number of head 
rows in class. 


Range in percentage of 
hard grains. 


Average yield per 100 
feet of row. 


Natural ten- 
dency under 
Arizona con- 
ditions. 


Hard. 


Soft. 


Hard. 


Soft. 


Hard. 


Soft. 








Percent. 


Percent. 


Grams. 


Grams. 




White durum .... 


95 


169 


5-76 • 


77-100 


1,896 


2,108 


Hard 


Red durum 


14 


23 


18-70 


74-100 


1.434 


1,510 


do. 


Poulard 


48 


31 


0-27 


31- 92 


1,686 


1,560 


Soft 


White bread 


79 


IOO 


0-52 


53-ioo 


1,788 


1,283 


do. 


Red bread 


5i 


14 


4-21 


27-100 


3,4io 


2,111 


do. 



26 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



head row contained. In this table the range in percentage and 
average percentage of hard kernels in each group are given, with the 
average yield from 100 feet of row. 

Table I shows that in the white and red durum varieties there is a 
plus correlation between yield and hardness, whereas in the poulard 
and white and red bread wheats the correlation between yield and 
hardness has the minus sign. In other words, among the durum 
wheats the harder strains on the average produced most ; whereas 
among the poulards and bread wheats the high yielders were softest. 
As it has been repeatedly shown (Lyon and Keyser, 3 Howard, 
Leake, and Howard 4 and Le Clerc and Leavitt 5 ) that the soft, opaque 
yellow berries contain less protein than the hard, horny, translucent 
grains of the same variety, we may safely assume that these harder 
grained strains also contained a higher nitrogen or protein content 
than those with a larger percentage of soft kernels. Now, since 
practically all commercial varieties of wheat are mixtures of slightly 
different types and since under the given conditions certain strains 
produce more than others, when these varieties are sown in a given 
locality for a number of years, a climatic selection will commence 
whereby the higher yielding strains will gradually come into ascend- 
ency. If now these high yielding strains be the softer ones, the 
harder strains will be slowly eliminated and we shall find the wheat 
becoming softer. 

One hundred and forty-five head selections were made from a com- 
mercial variety of Turkey wheat obtained in Kansas. All of these 
selections were true to the Turkey type of head, habit of growth, and 
lize and character of kernel. The pure races which they produced, 
however, differed considerably in their yield and in their resistance 
to the tendency to become soft under Arizona conditions. Thus the 
percentage of hard grains in the different strains in 1914 varied from 
3 to IOO percenl ; ill [915, from 29 tp EOO percent, and in T916 from 
05 to 100 percent. 'I bat these differences were varietal and tended 
t" persist in the same strains from year to year is shown by the 
correlations between these characters from one year to the next, as 
follows : 

'Lyon, T. L, am! Keyser, A. Winter wheat Nebr, Afjr. Expt. Sta. Bui. 89. 
if/15 
4 Loc cit. 

• Lc Clcn ) v. gad Leavitt, S. Tri-local experiment! on the influence of 

cnviromnrnl on tin- physical and chemical characteristics of wheat. U. S. Dept. 

Aki*., Jour. Art. Research, v. 1, no. 4, p. 276-291. 1914. 



FREEMAN : HARD AND SOFT KERNELS IN WHEAT. 



27 



Percentage of hard grains in 1914 with 

percentage of hard grains in 1915 = 57% + 4% 
Percentage of hard grains in 1915 with 

percentage of hard grains in 1916 = 33% + 5% 
Yield in 1914 with yield in 1915 = 55% + 4% 
Yield in 1915 with yield in 1916 = 68% + 3% 

The correlation between yield and percentage of hard grains in these 
145 races for the three years were as follows: 

1914 — — 25% + 5% 

1915= — 17% ±5% 
1916 = — 25% + 5% 

The meaning of this may be illustrated by a short study of the results 
in 1914. The 145 strains may be divided into three approximately 
equal groups and the average percentage of hard grains and the yield 
of each group calculated as shown in Table 3. If these strains be- 
haved the same each year (as it has been shown that they do), it is 
easy to see that the harder races will be practically eliminated in a 
few years. 



Table 3. — Correlation between yield and percentage of hard grains in Turkey 

wheat in 1914. 



Number of 
races. 


Average percentage of 
hard grains. 


Average yield in grams per 
100 feet of row. 


Percentage of total yield. 


54 


52 


1,989 


53 


38 


59 


1,387 


27 


53 


73 


744 


20 



If, for simplicity of calculation, the races are divided into a harder 
and a softer group based upon their average percentage of hard 
grains for the three years and the results from mixed planting cal- 
culated on the basis of their respective yields the following figures 
are obtained. 



Year. 


Crop. 


Percentage 


Percentage 


in softer group. 


in harder group. 





Seed for 19 13 


52 


48 


I 


Crop of 1914 and seed for 1915 


55 


45 


2 


Crop of 191 5 and seed for 1916 


59 


4i 


3 


Crop of 1916 and seed for 1917 


65 


35 



Projecting the results from the average of the past 3 years : 



28 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Year. 


Crop. , 


Percentage 
in softer group. 


Percentage 
in harder group. 


4 


Crop of 191 7 and seed for 1918 


70 


30 


5 


Crop of 1918 and seed for 1919 


75 


25 


6 


Crop of 1919 and seed for 1920 


79 


21 


7 


Crop of 1920 and seed for 192 1 


83 ' 


17 


8 


Crop of 192 1 and seed for 1922 


86 


14 


9 


Crop of 1922 and seed for 1923 


89 


II 


10 


Crop of ig.'j and seed for 1024 


91 


9 



The above tabulations shows that the hard group is thus practically 
eliminated at the end of ten years. The gradual softening of an 
impure race of wheat can thus be explained as a climatic selection 
without the necessity of assuming any direct or accumulative influ- 
ence of the climate upon the hereditary substance itself. 

In like manner we can understand why the durum wheats remain 
hard under Arizona conditions, for here there is a plus correlation 
between yield and percentage of hard kernels (+22 percent). An 
hereditary distinction between the durum and Turkey wheat is thus 
brought to light in that the harder strains of Turkey wheat are much 
reduced in yield whereas in the durum wheats the harder strains are 
the better yielders. These hereditary distinctions, though not strik- 
ing in any one season, are sufficient to maintain the hardness of the 
durum wheat and slowly change the other toward the condition of 
softness and low nitrogen content usually found among bread wheats 
which have been grown for a number of years in a warm climate. 

To the agriculturist, the economic conclusion is evident. We must 
discard mixed commercial varieties and grow only pure races of 
wheat coming originally from a single plant. As it is practically 
impossible to prevent the mixing of varieties through the custom 
thrashers on the farm, it is highly important that the seed wheat of 
the community be maintained in its standard of purity through re- 
I ■ ated pedigree selection. This work should be done either by the 
StS ' or by reputable trained seed breeders and from these the farmer 
should renew bis ieed .'it teasl every four or five years. 



macintire: growth of sheep sorrel. 



29 



THE GROWTH OF SHEEP SORREL IN CALCAREOUS AND 
DOLOMITIC MEDIA. 1 

W. H. MacIntire. 

The presence of sheep sorrel (Rumex acetosella) under field con- 
ditions is generally conceded to indicate a distinct need of lime. The 
thriving growth of this weed on soil known to be poor in lime was 
responsible for the common belief that the plant grows best in an acid 
medium. However, it is now known that sorrel will flourish in soil 
that has been treated with lime ; hence the viewpoint, now held most 
generally, that the plant thrives in acid soils because of lack of com- 
petition from those plants which are more sensitive to a low content 
of the "alkali-earthy elements. Recent contributions have been made 
upon this subject by White 2 and by Pipal. 3 

As confirming the findings of the two articles just mentioned and 
also as demonstrating the parallel effects of limestone and dolomite, 
the following brief statement is offered, together with the illustra- 
tions in Plate 1. This work was carried out by Dr. J. I. Hardy, now 
of the Wyoming station, under the direction of the writer. It was a 
portion of a series of pot-culture experiments which were incor- 
porated in a thesis as a part requirement for the degree of Master of 
Science during the years 191 2-1 3 and 191 3-14. 

Twenty 8-inch clay pots were used in the work, ten containing lime- 
stone in different percentages and ten containing dolomite in corre- 
sponding amounts. The pots were twice treated with asphaltum 
paint in order practically to eliminate porosity, so that they could be 
imbedded in the ground and thus insure more uniform and more 
nearly normal temperature. 

Clean river sand and limestone or sand and dolomite constituted 
the only solid material in each pot. Each of the three materials was 
sifted through a i-mm. sieve. Those portions which passed through 

1 Contribution from the University of Tennessee Agricultural Experiment 
Station, Knoxville, Tenn. Received for publication September 10, 1917. 

2 White, J. W. Concerning the growth and composition of clover and sorrel 
(Rumex acetosella) as influenced by varied amounts of limestone. Ann. Rpt. 
Pa. Agr. Expt. Sta., 1913-14, p. 46-64. 1914- 

3 Pipal, F. J. Red sorrel and its control. Ind. (Purdue Univ.) Expt. Sta. 
Bui. 197. 1916. 



30 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

the I -mm. sieve but were stopped by the y 2 -mm. sieve, were taken 
for the filling of the pots. The media varied from 100 percent lime- 
stone or dolomite to none, the difference between the percentage of 
each stone and 100 percent being represented by sand. The lime- 
stone and dolomite percentages were 100, 75, 50, 25, 15, 2.5, 1.0, 0.5, 
and o, as shown in Table 1. The densities of the sand and limestones 
were closely approximate and the aggregate weight and volume of 
the contents of each pot were practically constant. 

The limestone used contained 92 percent CaCO^ and 0.5 percent 
MgCO ;; , making a lime-magnesia ratio of 184 to I, while the dolomite 
contained approximately 50 percent CaC0 3 and 35 percent of MgC0 3 , 
or a ratio of 10 to 7. Each pot was supplied with 0.1 percent P 2 5 
in the form of precipitated lime phosphate and thoroughly mixed in 
the dry. No iron was added, since each of the stones carried im- 
purities of this element. A nutrient stock solution containing potas- 
sium sulfate, potassium nitrate, and potassium chloride was used as 
a source of mineral plant food. Aliquots of this solution were diluted 
and applied at the time of seeding and at subsequent intervals, ac- 
cording to the needs of the plants. In the two pots containing sand 
without limestone or dolomite there was of course some calcium 
present as phosphate and probably some traces in the sand itself; 
hence, 0.2 percent of precipitated MgC0 3 was applied to each of 
these pots in order that no one of the pots should have lime without 
magnesium. 

Sorrel seed was sown in the limestone pots April 2, 1913. The 
Seed was from the ammonium sulfate plats of the Pennsylvania sta- 
tion and was obtained through the courtesy of Prof. C. F. Noll, of 
thai station. A uniform stand was obtained in each of the limestone 
; otSj '»ut on June 13 the plants had grown only to a height of about 
an eighth of an inch . though they were apparently in normal con- 
dition. In older tO bring the growing period within the desired limits, 
the method of obtaining a stand was modified. The. seedlings were 
removed .and eight stolon^ of equal size were planted in each pot. The 

plant 1 were harvested September jj, 1913, thus giving a growing 

period of 101 flays. I he -ante procedure was carried out with the 
dolomite pots, the stolons being planted March 31 and harvested July 
6, I91 \. giving a growing period of <v> days. The yields from the 
I pol containing both limestones are given in Table 1 and the 
plants themselves arc shown in Plate 1. 

It will be noted thai the sorrel in the limestone pols was harvested 

tore the forming of teed, while fructification look- place in the dolo- 



Journal of the American Society of Agronomy. 



Plate 1. 




Fig. i. Growth of sorrel in limestone pots with constant lime-magnesia ratio 
of 184 to 1. From left to right, the percentage of limestone is 100, 75, 50, 25, 
I 5» 5> 2 > 5, 1, 0.5, and 0, the remainder in each case being sand. 




Fig. 2. Growth of sorrel in dolomite pots with a constant lime-magnesia ratio 
of 10 to 7. From left to right, the percentages of dolomite are 100, 75. 50, 2 5, 
15, 5, 2.5. 1, 0.5. and o, the remainder in each case being sand. 



THE LIBRARY 
OF THE 
UNIVERSITY OF ILLINOIS 



macintire: growth of sheep sorrel. 



3 1 



mite pots. However, the limestone series was planted and harvested 
in 1913, while the dolomite series was handled in 1914. In this 
respect the two sets of pots are not in strict comparison. It is there- 
fore inadvisable to make deductions as to the influence of the CaO- 
MgO ratios. 



Table i. — Air-dry weight in grams of entire plants of sorrel grown in pots 
containing varying percentages of limestone and dolomite. 



Pot No. 


Limestone series. 


Dolomite series. 


Limestone. 


Sand. 


Weight of 
plants. 


Dolomite. 


Sand. 


Weight of 
plants. 




Percent. 


Percent. 


Grams. 


Percent. 


Percent. 


Grams. 


I 


100.0 





1.61 


100.0 





3-34 


2 


75-0 


25.0 


.42 


75-o 


25.0 


6-54 


3 


50.0 


50.0 


1.06 


50.0 


50.0 


1.20 


4 


25.0 


75-0 


1.91 


25.0 


75-0 


i-5i 


5 


15.0 


85.0 


•75 


15.0 


85.0 


1.38 


6 


5-o 


95-0 


1.83 


5-o 


95-0 


1.02 


7 


2.5. 


97-5 


1.48 


2-5 


97-5 


i-57 


8 


1.0 


99.0 


1.24 


1.0 


99.0 


3 96 


9 


•5 


99-5 


•5i 


•5 


99-5 


3-87 


10 





100.0 


1.82 





100.0 


1.60 


Total 






10 81 






24-39 













The results secured demonstrate that the sorrel has no difficulty 
in maintaining a good growth in strongly alkaline media when not 
subjected to the intervening influence of clover or other lime-loving 
plants. 

It would seem that the conditions to which the growing plants were 
subjected in the limestone and dolomite pots are as severe as could 
be obtained. On the other hand, it is possible that more of the earthy 
bases might be offered in solution to the plant roots in the case of a 
fertile soil containing less of lime and more of organic matter ; that 
is, one not so rich in bases but richer in dissolved C0 2 . However, 
this may be in turn offset by the extensive area of contact between 
roots and limestone the solubility of which is appreciable even in dis- 
tilled water. As a matter of fact, it was impossible to separate 
mechanically the roots of the plants from the limestones to an extent 
which would permit the chemical analysis of the plants as a whole. 
The heavy root development shown in Plate 1 demonstrates very 
forcibly that an abundance of the earthy alkali carbonates is in no way 
inhibitory to the subsurface development of Rumex acetosella. 



32 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



VARIATIONS IN THE DEVELOPMENT OF SECONDARY 
ROOTLETS IN CEREALS. 1 

E. H. Walworth and L. H. Smith. 

In an article appearing in a previous number of this Journal, 2 it 
was shown that the number of temporary rootlets in cereals, instead 
of being constant within a given variety, is variable. For example, 
the author points out that instead of there being a whorl of three tem- 
porary rootlets in wheat, according to the usual description given in 
the literature, the number may be more or less than three and as 
many as five were noted in some instances. This inconstancy was 
likewise found to hold true with corn. 

Inasmuch as the present writers had an investigation along this 
line already under way when the above mentioned article appeared 
and have obtained additional data on this subject, the following ob- 
servations are presented with the thought that some further infor- 
mation along this line may be of interest. The results are reported 
of experiments with certain varieties of small grains, including oats, 
wheat, and barley, with respect to the variations existing in the num- 
ber of secondary rootlets. 

The term u secondary rootlet " is here applied to the temporary roots 
of the seedling other than the radicle and hence the latter is not in- 
cluded in the data presented. The method followed was to take 
representative samples consisting of coo or more kernels from each 
tot and sow them in pure quartz sand in the greenhouse. The 
counts were made when the plumules had attained the length of from 
i to inches. The results are set forth in Tables i and 2. Table I 
includes a li-t of sonic miscellaneous varieties, while Table 2 is made 
np of several selected -trains of wheat and of oats. 

In genera) our observations confirm those of Wiggans, referred to 
above, ifl that the number of secondary rootlets is by no means con- 
stant for a given variety, but varies among the individuals so that 

1 ( ontribution from the Haul Breeding Division of the Department of 
AKronomy, L'nivci it) oi Illinois, Urbana, III. Received for publication Octo- 
ber 13, 1917 

K'o-. f< Tin- number of temporary roots in cereals. In Jour. 

AriHT. So* A^ron., v K, no, 1, p. 31-37. 19,16. 



WALWORTH & SMITH: SECONDARY ROOTLETS IN CEREALS. 33 



Table i. — Variability of secondary rootlets in different varieties of the small 

grains. 

OATS. 



Variety. 


Number 
sown. 


Number 
nated. 




IS 


umber of 

2 3 


secoi 
4 


ldaiy 

5 


root 


lets. 
7 


Mean 


Wisconsin Pedigree No. 1 


100 


89 




3 


00 


2 5 


I 








2-3 


White Russian 


100 


82 


1 


4 


Oo 






I 








2.0 


Mammoth Cluster . 


100 


8l 


1 


10 


55 


12 


3 








2.1 


Schoenen 


100 


02 
y * 




1 


50 


38 


6 








2 S 


Garton's Victor 


200 


190 




6 


160 


23 


I 








2.1 


White Bonanza 


100 


QI 

y A 




A 


48 


38 


I 








2 A 


Silvermine 


100 


02 
y * 




1 



- T 


37 


I 








2 A 


Iowa No 105 


200 


176 




c 


55 


115 


T 








2.6 


Bi°" Four 


100 


QO 

y w 






54 


35 










2 4 


Silver Plume 


100 


77 




5 


58 


12 


2 








2.1 




100 


88 




5 


74 


8 


I 








2.1 




200 


191 




4 


41 


116 


29 


I 






2.9 


Victory 


2 00 


172 




4 


30 


115 


21 


2 






2-9 


Lincoln 


100 


95 




1 


24 


56 


14 








2.9 


Swedish Select 


100 


97 




1 


33 


54 


9 








2.7 




100 


98 




1 


46 


4i 


10 








2.6 




100 


9i 




1 


40 


47 


2 


I 




2.6 


Danish White 


100 


9i 




1 


60 


28 


2 








2.4 


Minnesota No. 6 


100 


94 






67 


26 


1 








2.3 


Great American 


100 


95 




3 


26 


65 


1 








2.7 


Sixtv Dav 


200 


192 




11 


121 


58 


2 








2.3 



WHEAT. 



Red Wave 


200 




| 1 


67 


28 


10 


1 






25 




200 


131 


j 7 


77 


32 


15 








2.4 


Turkev Hybrid 509 


300 


122 




25 


3i 


61 


5 






2.4 


Pesterboden 


300 


220 


4 


126 


47 


43 








2.6 


Dawson's Golden Chaff 


400 


179 




76 


49 


33 


21 






2.9 




500 


419 


9 


347 


•49 


14 








2.2 


K. B. 2 


200 


76 


1 


58 


11 


6 








2.3 


Red Hussar 


200 


129 




95 


25 


9 








2.3 


Turkey Hybrid 402 


200 


90 


2 


47 


29 


12 








2.6 


Red Cross 


500 


388 


9 


312 


55 


12 








2.2 


Durum 


200 


83 




9 


13 


60 


1 






3-6 



BARLEY. 



Oderbrucker 


200 


147 






1 


16 


55 


65 


10 


4-5 


Wisconsin Pedigree 


200 


152 






3 


14 


56 


66 


12 1 


4-5 


Two-rowed 


200 


119 




3 


8 


23 


46 


3i 


8 


4.0 


Beardless 


200 


114 






1 


15 


47 


45 


5 1 


4-4 



counts made on a random sample usually give a frequency distribu- 
tion represented by a fairly normal curve. 3 

In the case of the oats examined, the number of secondary root- 

3 It should be borne in mind that in order to compare our results with those 
of Wiggans, it is necessary to deduct one from the count in each case to allow 
for the radicle which was not included in our counts. 



34 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Table 2. — Variability of secondary rootlets in pure lines of small grains. 

OATS — PURE LINES. 





Number 


Number 


Number of 


secondary rootlets. 




sown. 


germinated. 










Mean. 






1 


2 


3 


4 


Silverminc 6-403 


IOO 


99 




42 


56 


I 


* 2.6 


Silvermine 14-383 


IOO 


92 


2 


24 


64 


2 


2-7 




IOO 


98 


I 


40 


55 


2 


2.6 


Black Gotham 13-332 


IOO 


97 




23 


74 




2.8 



WHEAT — INDIVIDUAL PLANTS. 



Malakoff 1 


70 


62 




61 


1 




2.0 


Malakoff 2 


100 


91 




90 


1 




2-3 


Malakoff 3 


IOO 


98 




95 


2 




2.0 


Malakoff 4 


90 


81 


1 


78 


2 




2.0 


Turkey Hybrid 509 — 1 


60 


55 




37 


13 


5 


2.4 


Turkey Hybrid 509 — 2 


IOO 


99 




7i 


20 


8 


?.4 


Turkey Hybrid 509 — 3 


IOO 


99 




62 


24 


13 


2.6 


Turkey Hybrid 509 — 4 


IOO 


74 




32 


21 


21 


2.9 



lets ranged from to 5. The table also brings out the fact that the 
different varieties have their characteristic tendencies toward a higher 
or lower number as shown by the column of averages. This dif- 
ference is also very well expressed by the number of greatest fre- 
quency, or modal number, which in some varieties is 2 while in others 
it is 3. 

Of the eleven varieties of wheat observed nine of them have the 
modal number at 2 while in the other two varieties 4 is the number 
of greatest frequency. This is a rather sharp distinction and it would 
be highly interesting to know whether any significant correlations 
exi^t between t hi > rootlet development and other characteristics of 
the wheal plant. We hope to pursue this matter in subsequent studies. 

The barleys exhibil a tendency toward a higher number of sec- 
ondary rootlets than possessed by either oats or wheat. Here the 
maximum numbe r rises to 7, while the modal number varied from 3 
in some varieties to 4 in others. 

It i| interesting to note that in the pure line selections of oats the 
range of variability 1- less than thai shown in the ordinary varieties, 
a result which perhaps might have been anticipated. This also holds 

i 1 tor single plants of wheat, although among the latter those of 
Turkey Hybrid 509 are slightly more variable than the others, which 
may possibly be accounted for by their hybrid origin. 

In connection with this study ;i somewhat more extensive series of 
Observations was made in maize Where an attempt was made to ascer- 



call & sewell: weed growth and nitrogen accumulation. 35 

tain whether any correlation exists between the development of these 
secondary seminal rootlets and the yielding capacity. We hope to 
have the opportunity of presenting some of these results in a future 
report. 

Summary. 

A study of the development of the secondary seminal rootlets in 
cereals has shown that : 

1. The number is not constant, but fluctuates greatly among indi- 
viduals, variations having been found ranging from o to 7. 

2. Different varieties of a given cereal show characteristic tend- 
encies in the production of these rootlets. 

3. As among the different cereals observed this tendency is greater 
in barley than in either wheat or oats, as indicated by the varietal 
averages, modal numbers, and highest extremes. 

THE RELATION OF WEED GROWTH TO NITRIC NITROGEN 
ACCUMULATION IN THE SOIL 1 

L. E. Call and M. C. Sewell 

iNTROl'UCTION. 

The accepted theory of the effect of tillage upon nitrification has 
been that tillage increased this biochemical action through a favorable 
influence upon the principal factors governing nitrification. These 
factors are the incorporation of organic matter, the distribution of 
bacterial flora, aeration, and moisture. The writers hope to show 
in this paper that in the past too much emphasis has been placed on 
tillage as an agent directly contributing to the formation of nitrates 
through its effect on the above factors and too little emphasis on it. 
as an indirect means of assisting in the accumulation of nitrates by 
preventing weeds from using them in their growth. 

Review of Literature. 

Schlosing and Miintz, Warington, Deherain (6), 2 and Marquenne 
(19) were among the earliest investigators to point out the relation 

1 Contribution from the laboratories of the Department of Agronomy (Paper 
No. 14), Kansas Agricultural Experiment Station, Manhattan, Kans. This 
paper embodies some of the results obtained in the prosecution of Project No. 
18 of the Kansas station. Presented by the senior author at the tenth annual 
meeting of the American Society of Agronomy, Washington, D. C, November 
12, 1917. 

2 Figures in parentheses refer to " Literature cited," p. 43. 



36 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



between nitrification and tillage through the incorporation of organic 
matter in the soil. 

The beneficial effect of barnyard manure on crops is common 
knowledge. That barnyard manure increases nitrification is shown 
by the recent work of Brown (i), who found that applications of 
manure up to 16 tons to the acre increased the number of organisms 
and the ammonifying and nitrifying powers of the soil. He also 
pointed out the relationship between bacterial activities and crop 
yields. Temple (18), two years prior to this, published similar re- 
sults showing an increase of bacterial activities as a result of appli- 
cations of barnyard manure. 

Schlbsing (6) advanced the theory that stirring the soil favors the 
spread of organisms. There are no data today that refute it. We 
know, however, that the nitrifying bacteria are widely scattered. 
Miintz and Aubin (6) observed them in many cultivated soils as well 
as in those of deserts and of high altitudes. 

Aeration. 

The value of tillage in aerating soils has been recognized since 
the time of Liebig (14), who pointed out that the decay of organic 
matter can only take place with a plentiful supply of oxygen, 
although putrefaction may occur with limited amounts of oxygen 
present. 

Schreiner and Sullivan (17) explained the important role of oxi- 
dation in the transformation of mineral and organic matter. They 
pointed out that whatever decreases oxidation in soils tends also to 
bring about the conditions which decrease growth, and the factors 
which favor oxidation arc the factors which favor soil productivity. 

The fact that oxygen is necessary for nitrification does not neces- 
sarily imply that an increase in the supply of oxygen will increase 
nitrification. Leather (11), as a result of recent investigations of 
:li<- -oil states that "it is certain that the diffusion of gases 

through -nil* at a depth of u to 15 inches is so efficient as to warrant 
the conclusion that cultivation of the surface soil is unnecessary for 
purposes of aeration." Mis investigations showed that even during 
tlx- wettesl weather the volume of ,^as falls only to [5 or 20 percent 
of the total soil volume or about half thai which is present during 
long periods of hot, dry weather. 

kusscll and \pplcyard (16) reported results showing but little 
variation in the composition of atmospheric and soil air. 

King and WhltSOfl (10) have presented investigations on the effect 



call & sewell: weed growth and nitrogen accumulation. 37 

of increasing aeration on nitrification. They bored holes in the soil 
and determined nitric nitrogen in the surrounding area. There were 
no data obtained which indicated that nitrification was increased by 
this manner of aerating the soil. 

In a paper on "The Soil Mulch" Call and Sewell (3) showed that 
nitrification is as great in uncultivated soil (silt loam) free of weeds 
as in cultivated soil of the same type. 

Gainey and Metzler (8) studied the rate of nitrification in a com- 
pacted and an uncompacted soil in the laboratory and found greater 
nitrification in the compacted soil up to the point where the moisture 
content reached two thirds saturation. 

These results show that while oxygen is essential to nitrification, yet 
sufficient aeration of the soil takes place under most conditions for 
optimum nitrification without cultivation. 

Moisture. 

Schlosing (6) observed that the nitrates formed in a soil increased 
with the moisture content until the moisture was sufficient to inter- 
fere with the free passage of air. 

Patterson and Scott (15) investigated the influence of soil mois- 
ture upon nitrification and concluded that — 

1. Nitrification is inactive in soil containing three times more moisture than 

in its natural air dry condition. 

2. For both sandy and clayey soils, optimum amounts of water for nitrification 

lie within the range of 14 to 18 parts per 100 of dry soil. 

Coleman (5) found that nitrification in a loam was most active in 
the presence of 16 percent of water and was much retarded when the 
amount of water was reduced to 10 percent or increased to 26 percent. 

Lill (12), working with the Marshall silt loam soil, found nitrifica- 
tion active between limits of 5 to 35 percent of moisture. He found 
two maxima, one at 15 percent and one at 30 percent moisture con- 
tent. 

Hutchinson (9) asserts that in the Pusa soil the optimum moisture 
content for nitrification is 16 percent and that general bacterial action 
is intense up to 25 per cent. 

Gainey (7), working with a silt loam soil, found an increase of 
nitrification in loose soil with an increase in moisture up to 30 percent. 

The writers have shown in a previous paper (3) that the moisture 
content of uncultivated soil free of weeds equals that of cultivated 
soils. Thus cultivation could only effect moisture as a factor in- 
fluencing nitrification through the control of weed growth. 



38 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

The results of these investigations indicate that the amount of soil 
moisture has an important bearing on the rate of nitrification and that 
the moisture condition most favorable for nitrification is about the 
same as the optimum condition for plant growth. It is also shown 
that nitrification may be decidedly active with a moisture content 
much below the optimum. Some discrepancy in the results reported 
by different investigations is to be expected, as widely different soils 
were used. 

A review of these investigations as a whole indicate that tillage in 
its effect on the factors influencing nitrification, particularly in its 
effect on aeration and moisture, will not account for the difference 
in nitrates found to exist in farm practice in soil cultivated in dif- 
ferent ways. 

Experimental Results. 

That a marked difference in the nitrate content of soil cultivated 
in different ways does occur in farm practice is shown by experiments 
conducted at the Kansas station in which soil has been prepared for 
wheat in eleven different ways through a period of nine years and in 
which the moisture and nitrate content of the soil was studied (2). 
In this experiment wheat was grown continuously and each plot re- 
ceived the same preparation each year. The average yields, the per- 
centage of water available for growth at seeding time, and the amount 
of nitrates in the soil at seeding time in each plot are given in Table 1. 



Table i. — Seven-year average yield of wheat, moisture content, and nitrates in 
parts per million to a depth of 3 feet at the time of fall seeding 
on 11 plats variously treated. 



Methods of preparation. 


Yield per acre. 


Moisture 
available to 
plant growth. 


Nitrates to a 
depth of 3 feet. 




Bushels. 


Percent. 


P. p.m. 


I Ji-k«-f I at seeding 


7-7 


4.8 


6-7 , 


Plowed Sept. 1 5, 3 inches deep 


I3-S 


5-6 


9-5 


Plowed Sept. 15. 7 inches deep 


14.8 


6.2 


7.0 


[ Jinked July i s. plowed Sept. 15, 7 inches deep 


19.2 


6.0 


17.6 


Plowed Auk. i S. 7 inches (Jeep, and worked as 




needed 


20.7 


6.8 


17.4 


Plowed Auk- i S. 7 indie* deep, not worked un- 






til Sept 15 


IQ.I 


6.6 


16.6 


Dmked July is. plowed Auk- I S. 7 indies deep 


I9.4 


6.2 


24.2 


Plowed July 1 5, 7 indies (Jeep, and worked as 






needed 


22.0 


5.1 


2 s.i 


Plowed July 1 5. \ indies deep 


17. 1 


5-7 


20.7 


Liitrd July 15, ridges worked down 


I8.5 


6.3 


83.4 


Uwtrd July is. ridges split Aug, 15 


IH.2 


5-8 


21 .1 



The amount of nitrates in the -oil at seeding tinif and the subse- 



CALL & SEWELL : WEED GROWTH AND NITROGEN ACCUMULATION. 39 



quent yield of wheat were much higher in all cases for early plowing 
or early preparation of the ground. The large difference in yield 
can not be attributed to a difference in moisture because the moisture 
content for all plots is nearly the same. There can scarcely be a 
doubt of the relation between the development of nitrates and yield 
of wheat. The question then arises of how to account for this large 
increase in nitrification and yield with early preparation if we accept 
the conclusions previously presented. 

Some additional data bearing on this question have been obtained 
by the writers in a study of the effect of weeds and of different 
depths of cultivation on moisture and nitrate accumulations in the 
soil. In this study there were four plots ; one was cultivated 3 
inches deep, one 6 inches deep, one was uncultivated but the weeds 
were removed with a hoe or by hand, and one was uncultivated and 
the weeds allowed to grow. 

As much moisture was found in the plots that were kept free of 
weeds but not cultivated, taking a 4-year average, as in the plots 
cultivated 3 or 6 inches deep (3). The nitrates expressed as pounds 
per acre 3 feet are shown in Table 2. These data represent the 
average of samples taken monthly from April to October each year. 



Table 2. — Annual and average development of nitrates in plots variously treated, 
expressed in pounds of A 7 3 per acre in the upper 3 feet of soil, in 
the four years from 1914 to 1917, inclusive. 



Treatment. 


1914. 




1916. 


1917. 


Average. 

1 

1914-16. 1914-17- 




124.3 


42.1 


78.5 


75-0 


81.6 


79-9 


3-inch mulch 


497-8 


495-9 


246.3 


225.3 


413-3 


366.4 


6-inch mulch 


550.1 


325-2 


567.8 


No data 


481.O 




Bare surface 


712.7 


643.0 


313-3 


228.9 


556.3 1 


4*74-5 



It is shown that more nitrates were developed in the uncultivated 
soil kept free of weeds than in the soil cultivated either 3 or 6 inches 
deep. It is also shown that there is much less nitrate in the soil 
which produced a growth of weeds than in the soil which did not 
produce weeds. It is evident that weed growth and not lack of cul- 
tivation is the factor responsible for the low nitrate content of the 
soil which produced weeds. 

In 1 91 6 and 191 7 the quantity of nitrogen contained in the weeds 
grown on the weed plots was determined. The amount of nitrogen 
in the weeds was calculated as nitrates and added to the nitrates 
present in the soil. These data are assembled in Table 3 and are 



4 o 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



compared with the nitrate content of the cultivated and uncultivated 
(bare) plots. 



Table 3. — Nitrogen in weeds expressed as pounds of N0 3 per acre, plus nitrates 
in the surface 3 feet, of soil in October. 



Year. 


X0 3 in soil culti- 
vated 3 inches. 


NO3 in soil with 
bare surface. 


NO3 in soil pro- 
ducing weeds. 


J Total N0 3 devel- 
Nitrogen in weed ed in weeds 
growth as N0 3 . and soil. 


IOl6 

1917 


531-5 
372.0 


445-7 
361.2 


158.3 
36.5 


316.O 474-3 

322.3 ! 358.8 



The last column of Table 3 gives the total amount of nitric nitro- 
gen which must have been produced in the soil growing weeds, pro- 
viding the weeds assimilated all the nitrogen used in their growth 
in the nitrate form. 3 These amounts nearly equal the nitrates pres- 
ent in the cultivated or bare surface soils. The estimate of nitric 
nitrogen in the weed growth is only for that contained in the tops. 

Referring again to the plots prepared in different ways for wheat, 
it will be seen that similar results have been obtained in regard to 
the effect of weeds upon nitric nitrogen accumulation in the soil. 
Determinations of the amount of weed growth and their nitrogen 
content have been made for the past three years on Plot I, which 
is un worked throughout the summer and prepared for wheat by disk- 
ing each season just before the wheat is sown. The nitrate content 
of the soil of thi- plot is compared with that of Plot 9, which is 
plowed 7 inches deep in July and worked throughout the summer. 
These data are presented in Table 4. 



Tahi h 4.— Nitrogen expressed OS nitric nitrogen (pounds of NO») per acre in 
Weeds and in the surface 3 feet of soil, September determinations. 





Nitrates in the soil. 


Nitrogen ex- 
pressed us nitric 
nitrogen in 
weeds, plot 1. 




I difference in 
total nitrates 
in favor of 

pl()l 'I. 


Dm* 


Plot 1. 
disked at 
sccdinK. 


Plot 9, 
plowed in 
July. 


I >ifleren< e in 
favor of plot 9. 


Nitric nitrogen 
in plot 1 , m. hid 
ii>K I hut in weeds. 


1915 
igi6 
1017 


18.8 
96.2 

13 4 


127.4 
36J.O 
186.5 


IO8.6 
265.8 
I73I 


I5'> 
167 
258 


168.8 
263.2 
271.4 


-41 4 

98.8 
-84.9 



hi this c;tM- greater nitrification appears to have taken place in 
• ir - M/15 and 1917 upon I Mot 1 than upon Mot evetl though 
it is only the nitrogen in weed tops that has been considered. As an 

l» is known tli.it plants can assimilate nitrogen in the form of ammonia and 
m IiikIiI> organi/rd compounds, l,ul it is not generally believed that they secure 
murh nitrogen in this way under ordinary conditions 



call & sewell: weed growth and nitrogen accumulation. 41 

average for the three years Plot 1 shows a higher accumulation than 
Plot 9. 

Further evidence may be obtained from these plots during the 
season of 191 3, a season which was so dry that weeds failed to 
grow upon Plot 1. This plot is unworked throughout the summer. 
Table 5 shows the precipitation for July, August, and September and 
the nitric nitrogen in the soil September 13 and October 10. 



Table 5. — Nitrate (pounds of N0 3 ) per acre to a depth of 3 feet in 1913 on 
Plot 1 (disked Oct. 1) and on Plot 9 (plowed in July), with the 
precipitation in July, August, and September. 



Precipitation in inches. 


NO3 in soil September 13. 


NO3 in soil October 10. 


July. 


August. 


September. 


Plot 1. 


Plot 9. 


Plot 1. 


Plot 9. 


0.07 


0-37 


4.89 


92.9 


76.5 


255-5 


258.7 



The precipitation during July and August was very light and the 
rains in September occurred after the determinations were made on 
the 13th. It will be seen that the accumulations of nitric nitrogen in 
the two plots when the last determinations were made are practically 
the same. This would tend to show that the small amounts of ni- 
trates found in the soil of Plot i in other seasons (Table i) are due 
to the fact that the weeds have used this compound in their growth. 

Numerous cultivation experiments with corn show that the prin- 
cipal benefit of tillage is the removal of weeds. Cates and Cox (4) 
tabulated the results of 125 experiments carried on for six years, 
1 906-191 1, in 28 different States. They concluded that cultivation 
is not beneficial to the corn plant except in the removal of weeds. 
Mosier and Gustafson (16), as a result of eight years' work, showed 
that killing weeds without cultivation produced a gain of 17.1 percent 
or 6.7 bushels per acre over ordinary cultivation (shallow three 
times) . 

At the Kansas station similar data regarding the effect of tillage 
on corn yields are available for the past three years. These results 
are shown in Table 6. 



Table 6. — Annual and average yields of corn on fall-plowed land, variously 
cultivated, 1914 to 1916, inclusive. 





Yield per acre, bushels. 




Cultivation treatment. 






1916. 


Average. 




1914. 






13.O 


65.O 


43.2 


40.6 




13-4 


62.0 


43-2 


39-5 


Ordinary and i horse cultivator every 10 days. . 


£1.0 


58.8 


43-4 


37-7 




9.2 


65.O 


45-2 


39-8 



42 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

As an average of the three years the uncultivated plots where the 
weeds were removed produced practically as high yields as the cul- 
tivated plots. Apparently there was no advantage from the point 
of yield in cultivating corn except for the purpose of killing weeds. 

Gates and Cox in their report refer to the early work upon the 
weed factor in cultivation, citing results of Sturtevant at the New 
York State Station at Geneva in 1886, in Illinois by Morrow and 
Hunt in 1888 to 1893, in Missouri by Sanborn and Waters in 1889 
and 1890, and in South Carolina in 1898 and 1899. Decisive evi- 
dence in favor of cultivation was not secured at any of these stations. 

If the difference in nitric nitrogen accumulation in cultivated and 
uncultivated soils is due entirely to the weed gro'wth upon the latter, 
then the depth of tillage should not affect the development of nitrates. 
That the depth of cultivation has not greatly affected the accumula- 
tion of nitric nitrogen in the soil of the plots prepared in different 
For wheat is shown by the results of the wheat seed bed experi- 
ment already discussed in Table 1 and further summarized in 
Table 7. 



Table 7. — Average nitrates per acre in p. p. m. in the surface foot of soils 
variously treated during the periods of cultivation in the eight 
years from 1909 to 19 1 6. 





Nitrates as p. p.m. of NOj in the surface toot. 


I'reatnient. 












July. 


Aug. 


Sept. 


Oct. 




10.3 


10.3 


7-4 


13-7 


Double-dished la July, plowed in August 7 










inches deep 


28.9 


31-9 


50.6 


55-6 


Double-disked in July, plowed in September 










7 inches deep 


24.4 


34-1 


26.3 


42.6 


Plowed in August 7 inches deep 


18.3 


25-9 


39-5 


38.3 


Plowed in July 7 inches deep 


i8. S 


37-8 


48.3 


50.8 


Plowed in July \ inches deep 


18.8 


32.4 


36.8 


44-5 


P!«>w«-t| in September 7 inches deep 


23-3 


1 2.7 


8.0 


16.0 


Plowed in September < inches deep 


17.0 


1 7.0 


11. 7 


21.3 



I li<- -nil double disked in July, thus preventing weed growth, and 
plowed 7 inches deep one month later has an equal or even greater 
nitrate content than the soil plowed 7 inches deep in July. The soil 
double disked in July and plowed 7 inches in September contains a 
much greater nitrate content than the soil plowed 7 inches in Sep- 
tember without previous disking. The soil plowed 3 inches in Sep- 
tember contains a nitrate content equal lo thai plowed 7 inches at the 
• date 'I bis effect of depth of tillage is not entirely substan- 
tiated bj the data on plowing 7 and 3 inches deep in July, as the 



CALL & SEWELL : WEED GROWTH AND NITROGEN ACCUMULATION. 43 

former exceeded the latter in nitrate content. However, the aver- 
age carbon and nitrogen analyses of Plots 9 and 15 may explain this 
difference in nitrification. The carbon content of the surface 7 
inches of Plot 9 has been 1.74 percent; of Plot 15, 1.53 percent. 
The nitrogen content of Plot 9 has been 0.148 percent; of Plot 15, 
0.131 percent. Plot 9 then contained 14 percent more carbon and 
13 percent more nitrogen than Plot 15. It seems possible that this 
difference in the quantity of nitrates liberated in this case may be due 
to original differences in the soil upon which the work was done. 

Summary. 

It appears from the data presented that in the past too much em- 
phasis may have been placed on tillage as a direct means of conserv- 
ing moisture and liberating plant food and too little emphasis on 
it for the purpose of destroying weeds. If moisture is lost from the 
soil principally through weed growth and if nitrogen and other ele- 
ments of plant food become available rapidly in unstirred soil, it is 
a matter of economy to handle the soil so that weeds may be con- 
trolled with the minimum of labor. 

It should not be understood that tillage is unessential. It will be 
necessary to cultivate ground to maintain the proper structural con- 
ditions of the soil, to dispose of crop residue on the surface of the 
soil, to incorporate manures and organic matter in the soil, and to 
place the soil in suitable condition for seed. Further than this, with 
the possible exception of heavy types of soil, it is doubtful if tillage 
is essential where the soil is in a receptive condition to absorb rain- 
fall and where there is no weed growth. 

Systems of good farming should be practiced which will assist in 
controlling weeds, such as a rotation of crops, the use of livestock, 
especially sheep, for grazing purposes, and timely tillage. By the 
use of such methods cultivation may be reduced without a corre- 
sponding reduction in crop yields. 

Literature Cited. 

1. Brown, P. E. 

1913. Bacteriological effect of barnyard manure. Iowa Agr. Expt. Sta. 

Research Bui. 13. 

2. Call, L. E. 

1914. The effect of different methods of preparing a seed bed for winter 

wheat upon yields, soil moisture, and nitrates. In Jour. Amer. 
Soc. Agron., 6 : 249-260. 

3. , and Sewell, M. C. 

1917. The soil mulch. In Jour. Amer. Soc. Agron., 9: 49-6 1 - 



44 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

4. Gates. J. S., and Cox, H. R. 

1912. The weed factor in the cultivation of corn. U. S. Dept. Agr., Bur. 
Plant Indus. Bui. 257. 

5. Coleman, Leslie C. 

1908. Nitrification. Abs. in Jour. Chem. Soc. (London), v. 94, no. 546, 

II, p. 315. 

6. Deheraix. P. P. 

[894. Nitrification in arable soil. /// U. S. Dept. Agr., Expt. Sta. Rec- 
ord 6: 353-366, 49I-50I- 

In this article the author reviews the work of Schlosing, 
Mtintz, Aubin, Warington, and others on nitrification. 

7. Gaixev, P. L. 

1916. The effect of time and depth of cultivating a wheat seed bed upon 

bacterial activity in the soil. In Soil Science, 2 : 193-204. 

8. . and Metzler, L. F. 

1917. Some factors affecting nitrate nitrogen accumulation in the soil. 

/;/ U. S. Dept. Agr., Jour. Agr. Research, v. 11, no. 2, p. 43-64. 

9. Hutchixsox, C. M. 

1911-12. Report of the Imperial Agricultural Bacteriologist. In Rpt. 
Agr. Research Inst, and Col. Pusa [India], p. 80. 

10. Kixg, F. H., and Whitson, A. R. 

1902. Development and distribution of nitrates and other soluble salts 
in cultivated soils. Wis. Agr. Expt. Sta. Bui. 93. 

11. Leathkr, J. \Y. 

1915. Soil gases. In Mem. Dept. Agr. India. Chem. Ser., v. 4, no. 3, p. 
85-134. 

12. Lill, J. G. 

1910-11. The influence of moisture upon nitrification. Unpublished 
tlusis. Kans. State Agr. College, Dept. Agr. 

13. Liebig, Justus. 

1852. Liebig's Complete Works on Chemistry, p. 44. Edited by Lyon 
Playfair. T. B, Peterson, Philadelphia. 

14. Mosikr, J. G , and Gustafson, A. F. 

[915. Soil moisture and tillage of corn. 111. Agr. Expt. Sta. Bui. 181. 

15. Patterson, J. W., and SCOTT, P. R, 

[912. 'I he influence of soil moisture upon nitrification. In Jour. Dept. 

\ur. Victoria, v, io, no. 5, p. 275-282. 
id Russell, E. J., and Appleyard, A. 

1015 The atmosphere of the soil, its composition and the causes of 

variations. J.. nr. Agr. Sci. (England), v. 7, pt. 1 (1015). 

p. I-4H. 

17 s< hbiimu, Oswald, and Sullivan, M. X. 

1910 Studies in loil oxidation. I '. S. Dept. Agr., I'>ur. Soils Bui. 73- 
18. Tkmple, J. C 

loll l'...Mi'..o,d rii.miire. effect upon the bacterial flora of the soil. Ga. 
Agr. Kxpt. Sta. Bui. 95. 
V) Wakinmon, Kohkkt. 

1892 'lb' Kothamsted Kxpcrimenl Station lectures. /;/ U. S. Dept. 
\g» . Offin- of l-.xperimetil Stations I'.ul. H. 



HARTWELL & PEMBER! ALUMINUM AND ACID SOILS. 



45 



ALUMINUM AS A FACTOR INFLUENCING THE EFFECT OF 
ACID SOILS ON DIFFERENT CROPS. 1 

Burt L. Hartwell and F. R. Pember. 

Attention has been directed by various investigators to the in- 
jurious effects arising from the hydrolysis of aluminum salts be- 
cause of the free acid caused thereby as measured by an increase in 
the concentration of the hydrogen ions. The entire emphasis has been 
laid heretofore upon the increase in acidity as the disturbing factor 
and not upon the aluminum itself. 

The Rhode Island station has for a number of years been in- 
terested in attempts to ascertain why different kinds of plants varied 
so remarkably in their response to liming. 2 For example, under the 
same conditions barley may be increased two to three times by liming 
and rye receive no benefit whatever. Nevertheless, the authors found 
that the addition of acid to ordinary nutrient solutions had as depress- 
ing an effect upon rye as upon barley seedlings. 3 From this it seemed 
probable that the toxicity of so-called acid soils was not attributable 
only to the acid, for in that case the two seedlings should have been 
affected alike. 

Whenever the influence of the various factors which we have 
studied was the same on the two seedlings we have been disinclined 
to accept those factors as being very helpful in explaining the vary- 
ing needs of different crops for lime. 

The aqueous extract of an acid soil, like the soil itself, affected the 
two kinds of seedlings very differently, showing that it contained some 
ingredient not present in an ordinary nutrient solution. 

Sterilization, dialysis, partial distillation, etc.. indicated that the 
substance was crystalloidal in nature. By evaporating the extract, 

1 Contribution No. 240 from the Agricultural Experiment Station of the 
Rhode Island State College, Kingston, R. I. Presented by the senior author at 
the tenth annual meeting of the American Society of Agronomy, Washington, 
D. C, November 13, 1917. 

2 Hartwell, Burt L., and Damon, S. C. The comparative effect on different 
kinds of plants of liming an acid soil. R. I. Agr. Expt. Sta. Bui. 160. 1914. 

3 Hartwell, Burt L., and Pember, F. R. The relation between the effects of 
liming, and of nutrient solutions containing different amounts of acids, upon 
the growth of different cereals. In 20th Ann. Rpt. R. I. Agr. Expt. Sta. 
(1906/07), part 2, p. 358-380. 1908. 



46 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



igniting, and dissolving the residue in acid, a culture medium was 
obtained which was much more injurious to barley than to rye. 

Aside from the ordinary nutrients the residue was found to con- 
tain silicon, aluminum, and chromium ; titanium and certain other 
ingredients which may have been present were not sought. Owing 
to the presence of considerable aluminum, studies were begun to 
determine the specific effect of this element by itself and in combina- 
tion with silicon, chromium, and other substances. 

Although the work had to be interrupted frequently, the cumulative 
impression caused by the results led to a disinclination to accept 
acidity as the only, or perhaps the main toxic factor influencing the 
growth of plants on acid soils. 4 

When c.p. aluminum sulfate about equivalent to the amount of 
aluminum found in the soil extract was used in nutrient solutions the 
barley seedlings were depressed much more than the rye, but when 
the same amount of sulfuric acid unaccompanied by the aluminum 
was present the rye was depressed as much as the barley. 

The hydrolysis of the aluminum sulfate was sufficient to give 
only about one fourth the concentration of hydrogen ions as that 
resulting from the equivalent amount of free acid. It seems there- 
fore that aluminum must have been the main cause of the depression 
in the growth of barley and that its effect on rye was much less. In 
other words, the two seedlings were affected differently by the 
nutrient solution containing aluminum, the same as they were by the 
aqueous extract of an acid soil. 

If aluminum as such is an important factor to be considered in 
connection with the deleterious effect of acid soils, any treatment 
which renders it less active will prove beneficial. It seemed reason- 
able that a thorough treatment of the soil with a soluble phosphate 
mighl accomplish this object. A moist acid soil upon which most 
kinds of plants were unable to exist was kept intimately mixed for 
about two weeks with acid phosphate added at the extraordinary 
rate or 28 tOIU per acre, after which lettuce was planted. This crop 
COtsld not exist in the unphosphated BOil supplied only with nutrients, 
but tlx- soj] treated with the acid phosphate produced a maximum 
« wn mop- than when lime replaced the phosphate. 

It was shown that for a considerable time at least the large amount 

'Hartwell, Burl L, and Damon, S. C. Loc cit., p. 410. Also Hartwell, 

I- Twenty eighth annual report of the Director of the Agricultural Kx- 
prrimeni Station. In Report ..f the Hoard of Managers, Hul. K. I. State Col- 
lege, vol. 11, no. 4, p. 28. Feb., 1016. 



AGRONOMIC AFFAIRS. 



47 



of acid phosphate greatly increased the acidity, and yet a crop which 
usually responds markedly to liming had made its maximum growth 
on a very acid soil without the addition of any lime. The solubility 
of the aluminum in dilute acetic and carbonic acids had been markedly 
reduced by the phosphate, just as it doubtless would be by lime or 
by a mixture of the two. 

Determinations of the amount of what may be called active alumi- 
num may prove to be as desirable as acidity determinations, and the 
lime requirements of a soil may be due to the need for lime to pre- 
cipitate toxic aluminum quite as much as to neutralize soil acidity. In 
fact, the authors found that after sufficient hydrated lime had been 
added to produce a maximum crop of lettuce a lime requirement 
equivalent to from 4,000 to 5,000 pounds of calcium oxid per acre, 
according to a procedure yielding results similar to the Veitch method, 
still existed at the end of the vegetation experiment, in spite of the 
fact that nearly all the lime had entered into reaction with the soil. 
Many instances might be cited for the economic application of acid 
phosphate apparently in excess of its need for nutrient purposes. 

The details of the work upon which are based the ideas here pre- 
sented will be published in other connections, but it is hoped at this 
time to interest agronomists in the application of the results to their 
problems. 

AGRONOMIC AFFAIRS. 
MEMBERSHIP CHANGES. 

The membership reported in the November number was 653. Since 
that time 6 new members have been added and 1 has been reinstated, 
while 6 have resigned and 1 not previously reported has been dropped 
for -nonpayment of 1916 dues. The membership therefore remains 
at 653. The names and addresses of the new members and of the 
reinstated member, the names of those who have resigned or lapsed, 
and such changes of address as have come to the notice of the Secre- 
tary are given below. 

New Members. 

Alexander, L. L., Farm Crops Dept., College of Agr.. Columbia, Mo. 
Freeman, H. A., Central Expt. Farm. Ottawa, Ontario, Canada. 
French, W. L., 1221 Laramie St., Manhattan, Kans. 
Hagy, F. S., 924 Fremont St., Manhattan, Kans. 
Kan, T. T.. Box 95, College Station, Texas. 

Stadler, L. J., Farm Crops Dept., College of Agr., Columbia, Mo. 

Member Reinstated. 
Garren, Geo. M., College of Agriculture, Raleigh, N. C. 



48 



AGRONOMIC AFFAIRS. 



Johx B. Abbott, 
Elmer D. Ball, 



Members Resigned. 
James M. Bell, 
Geo. F. Corson, 



Martin Nelson, 
M. H. Young. 



Member Lapsed. 
Jens Olsen. 



Changes of Address. 



Babcock, F. Ray, Crosby, N. Dak. 
Bixford, E. E., Stephenville, Texas. 
Chapman, James E., Granada, Minn. 
Crox. A. B., Box 1214, Amarillo, Texas. 

Dougall, Robert, The Davenport, 17 Kellogg Ave., Amherst, Mass. 

du Buissox. J. P., University of Stellenbosch, Stellenbosch, South Africa. 

Frank. W. L., Bureau of Markets, U. S. Dept. Agr., Washington, D. C. 

Hill, W. H.. Lai). Inland Rev. Dept., 249 Hastings St., E., Vancouver, B. C. 

Macfarlaxe, Wallace, 55 S Street, Salt Lake City, Utah. 

Mver, D. S.. Extension Div., Purdue Univ., La Fayette, Ind. 

Xevix. L. B., 433 Seventh St., Hollister, Cal. 

Packard, Walter E., Extension Div., Agr. College, Berkeley, Cal. 

Pfxi-lktox, Robert L., Agr. Dept., Ewing Christian College, Allahabad, India. 

Shantz. H. L., Chula Vista, Cal. 

Stokes. W. E., County Agent, Edgefield, S. C. 

Taggart, J. G.. School of Agr., Olds, Alta., Canada. 

VbOKHEES, John H., Agr. Lime Bureau, 503 Riggs Bldg., Washington, D. C. 
Wheeler, H. J.. 111 Grant Avenue, Newton Centre, Mass. 
Wl att. F. A., 216 Agr. Bldg., Urbana, 111. 



C. II. Bailey lias resumed his work in cereal chemistry at the Min- 
station after a year's leave of ahsence, during which he was in 
charge of the laboratory of the Minnesota Grain Inspection Depart- 
ment at Minneapolis. 

Ross L. Bancroft has been advanced from assistant professor to 
associate professor of s () j] fertility at Iowa State College. 

T. I'. < ooper, for the past several years dean of the North Dakota 
( ollege <>f Agriculture and director of the station, has been elected to 
a similar position in the University of Kentucky and entered on his 
dtStiei there January 1. 

I ! . ( haptnafl ha- been appointed instructor in soils at the Minne- 
1 »t;i 1 < ,!]' ■.'<■ and station. 

Walter k. Clark, of tin - department of political science of the Col- 

of the ( ity of New York, has been appointed presidenl of the 
University of Nevada and has entered 011 his new duties. 



NOTES AND NEWS. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. io. February, 1918. No. 2. 



CROP CENTERS OF THE UNITED STATES. 1 

Adolph E. Waller. 2 

The Relation of Vegetation to Evaporation and Rainfall. 

The geographic distribution of our important crop plants appears 
on investigation to be in accord with the well-known centers of 
natural vegetation. Attention has been called (Transeau, 1905) 3 
to the separation and restriction of groups of plants to regions 
where the combination of factors most suited to the development 
of the group was localized. Transeau was able to show this by a 
map of the rainfall evaporation ratios computed from data on evap- 
oration from a free water surface (Russell, 1888) and the known pre- 
cipitation for the same station. The ratio is an attempt to combine 
moisture and temperature data as related to plant growth into a 
single significant figure. His mapped results clearly indicate the 
desert region, the plains, the prairies and their eastern extension in 
Illinois, as well as the forest types of the East, namely, the central 
deciduous, the northeastern evergreen, the southeastern evergreen, 
and the insular tropical. Had Russell's evaporation data been more 
complete the ratios for the whole country could have been presented. 

As it is, the map (fig. 5) makes an acceptable working basis for 
outlining the vegetation of the North American continent and re- 

1 Contribution 99 from the Botanical Laboratory of the Ohio State Univer- 
sity, Columbus, Ohio. Received for publication August 7, 1917. 

2 The writer acknowledges with much pleasure the advice and assistance given 
him by Dr. E. X. Transeau during the preparation of this paper, which is a 
preliminary part of work being carried on under his direction. 

3 References are to " Literature cited," p. 81. 

49 



50 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 




waller: crop centers of the united states. 51 

mains still the best climatic chart that has been published on forest 
and prairie distribution. In spite of the discrepancy between rain- 
fall and the amount of water available for plants and between the 
evaporation and transpiration, the map serves as an effective means 
for visualizing the distribution of the vegetation. 

A number of attempts have been made to correlate plant growth 
with data gathered by the United States Weather ' Bureau. The 
evaporation studies made by this bureau are not, however, of much 
help to students of plant distribution, either agriculturists or ecolo- 
gists. Since evaporation and transpiration can be so conveniently 
used to summarize plant activities, we can go so far as to say that 
the data of the Weather Bureau are only in the most general way 
of importance to ecologists. 

To improve this situation, Livingston (1915), in a paper read 
during the Columbus meeting of the American Association for the 
Advancement of Science, suggested that the rainfall be measured 
in the usual way, but the evaporation be found by the use of 
standardized atmometers and the Lehanbauer method be used to find 
the index of temperature efficiency. By multiplying the rainfall- 
evaporation ratio by the physiological temperature index, a single 
figure representing the true climatic summation, namely, the moisture- 
temperature index, could be obtained. This suggested method also 
would avoid the inaccuracies of Transeau's chart. However, it 
has not yet become possible to obtain the results from this suggestion. 
The figures collected by Russell are still the most complete data 
available on the depth of evaporation. 

In measuring depth of evaporation three factors profoundly affect- 
ing plant life are involved : 

a, The temperature of the evaporating surface; 

b, The velocity of the wind; and 

c, The relative humidity. 

The rainfall-evaporation ratio, as it combines also the total pre- 
cipitation with those three, might be expected to agree closely with 
the actual occurrence of the known vegetational types of the country. 
This has been found to be the case. The similarities between fig- 
ure 5 and Sargent's (1884) map of the forests of North America 
are striking. 

A 110-percent rainfall-evaporation ratio, an amount that marks 
one of the boundaries of the northeastern evergreen forest, means 
that the evaporation called for is exceeded 1.1 times by the total 



52 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



annual precipitation. A 20-percent ratio means that the precipitation 
is only two tenths of the evaporation called for from a free evap- 
orating surface. Only plants with a marked ability, to conserve 
water can exist in the face of such severe conditions for plant 
growth as is found in regions where the rainfall is so inadequate. 

It is interesting to note the geographic significance and particularly 
the direction of the evaporation lines drawn through stations of the 
same ratio. In the Great Plains region in the vicinity of the 100th 
meridian a strong north and south tendency is evidenced. This indi- 
cates that in spite of the increase in temperature southward the ratio 
does not vary with the latitude. There is also an increase in pre- 
cipitation. The irregularities in the direction of the lines developing 
farther eastward are due to the proximity of large water areas, e. g., 
the Great Lakes, the Gulf of Mexico, and the Atlantic, and also to 
the direction of the winds bearing moisture from these reservoirs of 
the eastern United States. To a limited extent topography is also a 
factor. In charts showing the mean monthly and the annual rela- 
tive humidity, Johnson (1906) displays a distortion in the humidity 
lines similar to that which appears here in the rainfall-evaporation 
ratios. 

Climatic Origins. 

The antecedents of climatic variation are chiefly differences in the 
latitude, in the unequal warming of land and water areas, the 
elevation above sea level, and the direction and intensity of the pre- 
vailing winds as controlled by the occurrence and movement of 
anticyclones and cyclones. The differences in latitude classify those 
parts of the world which arc unlike with respect to the angle at 
which the- Min - raya strike the earth's surface. The amount of 
heal received and the length of the seasons depends upon the latitude. 
If tin- were the only eonsiderat ion and there were no surface factors 
to reaet on the atmosphere, it would be only a simple matter to 
divide the earth into climatic zones. 

The diffi rence in the specific heat of land and water, by which 
the rate of hea4 absorption and radiation of land and water areas 
is determined, is an important factor in establishing climate. This 
ence serves to divide climates into two principal classes, con- 
tinental and oceanic The sea takes up heat and gives it off again 
only one fourth as fast as the land. Climates that are influenced 
dominantly bj the sea or large bodies of water have moderate tem- 
changei between night and day and between winter and 



waller: crop centers of the united states. 



53 



summer. The oceanic type of climate is in other words equable. 
Inland, however, where the land absorbs heat and again radiates 
it three times faster than water does, the temperature changes be 
tween day and night and between winter and summer are relativel) 
rapid. Therefore, a variable or continental climate dominates. 
Since North America, next to Asia, contains the largest land mass, 
a vast territory in the United States lying between the Rocky 
Mountains and the iooth meridian possesses a severe continental 
climate. 

The greater the elevation above sea level the more rigorous the 
conditions for plant growth because of (a) lower temperature in 
summer and winter, (b) a drier atmosphere, accelerating evapora- 
tion, and (c) greater wind velocity. As compensation for these, 
there are two conditions favoring plant growth, (a) greater sun- 
shine intensity and, much more important, (b) greater rain and 
snowfall on the windward slopes. The heavier precipitation is due 
to the expansion of air as it is forced upward over the sides and 
summits of mountains in the paths of prevailing winds. On ex- 
panding, the air cools and the moisture it contains is condensed. 
Thus, the mountain sides receive more abundant downpours of 
rain than the adjacent lowlands. This effect of elevation is mark- 
edly evident in the Pacific Northwest coast region where the annual 
precipitation is more than 100 inches, the heaviest in the United 
States. The southeast slopes of the Appalachians also have a 
somewhat greater annual precipitation than the northwest slopes. 

The most noteworthy feature of the mountain ranges in America 
is their north and south trend across the paths of the prevailing 
winds. The lines of equal rainfall from the Rocky Mountains east- 
ward to the Great Plains are approximately north and south. In the 
Southern States east of Texas equal rainfall lines follow the gen- 
eral outlines of the Gulf Coast. In the Middle West and eastward 
to the Atlantic Ocean they are a complex of the combined effects 
of the Great Lakes, the Gulf, the Atlantic, and the reprecipitated 
moisture from the forests of eastern America. This last source of 
moisture for the prairies will be more fully discussed subsequently. 
The effect of the Appalachians on the rainfall of the territory lying 
between them and the source of water is much less apparent and 
also less direct than the effect of the western mountain systems. 

The next factors to be considered are the centers of action pro- 
duced by the large areas of permanent high and low barometric 



54 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



pressure. While all the permanent centers of action of the world 
have some effect on the climate of North America only five of these 
are shown in the accompanying diagram (fig. 6). For the purposes 




Fig. 6. Diagram of the permanent centers of influence of high and low baro- 
metric pressure in the vicinity of North America. The interplay of these and 
their effect upon the evaporating power of the air and on summer climatic con- 
ditions is indicated. 



of this paper it is sufficient to consider their operation in summer 
only, not in winter. Interaction of two kinds is to be considered. 
The first is the interaction between the anticyclones and the cyclones 
themselves. Ill addition, within one given area of action the opposite 
extremes have reciprocal effects. The eastern side of a high and 
the western side of a low are regions of ascending, converging, 
cooling air. This is a region of increasing moisture. The western 
tide of a high and the eastern side of a low are regions of descend- 
ing, diverging, warming, drying air. 

It will he noticed thai four of these permanent centers of action 
are at sea and only one on the continent. While they arc said to be 
permanent this doc. not mean thai they are stable. As a matter of 
tad they are constantly moving back and forth. If the continental 
low should shift ea-tward, or if the western side of the Atlantic 
high ihotlld encroach on the Atlantic coast, the effect on all the 
area <a t of the ir>oth meridian would he the same. Drying, clear- 
ing air would descend over the whole region, llowing either from 
i orthwett where the low exists, or from the northeast where the 
influence of the hi^h i main |Y -led. These two permanent centers 
of actSOfl are the moil important in their influence on summer 



waller: crop centers of the united states. 



55 



weather conditions of the greatest part of the United States. If 
the action of these centers should be just the converse of what has 
been described, i. e., if the Atlantic high did not encroach westward 
and the continental low remained in the far north, warm, moist 
weather conditions would prevail. The effect of these centers on the 
crops of eastern America is thus plain. If the drying, clearing air 
conditions lasted a long enough time, a drought and poor crops 
would result. If these conditions do not exist, then the prevailing 
winds are from the south and southwest. These, having blown 
over the reservoirs of the Gulf and the Atlantic, carry in moisture, 
which is precipitated as the air converges and ascends. 

The combined action of differences in latitude, difference in the spe- 
cific heat of land and water, elevation, and anticyclone and cyclone 
movements produce in the United States three distinct climates. The 
first is of the narrow strip of territory from the Pacific Coast to the 
mountains, an ocean type of climate purely. The second is of the 
upland plateau from the mountains eastward to the iooth meridian. 
Tn the intermountain basins the climatic conditions are locally modi- 
fied by the presence of water, especially northward, but over by far 
the greatest part of the region a changeable, continental climate 
dominates. The third division is from the iooth meridian where a 
continental climate prevails, to the Atlantic where conditions are 
influenced dominantly by the ocean. The change is gradual from one 
type to the other. 

On the coastal plains everywhere there are regions of rich, 
abundant vegetation. This is directly related to the climate, the 
vegetation being a response to the presence of large amounts of 
moisture. Climate is also the cause of the Great Plains, with their 
sparse vegetation and dry, shifting soils. The prairie may be re- 
garded as a transition from the forest conditions of the East to the 
arid conditions of the central plains. Climate is responsible for the 
origin of the prairies, but we must look to other factors to explain 
their persistence. The more abundant moisture of the prairies as 
compared with the plains has been attributed by Zon (1913), to 
the forests of the southeastern United States. In his opinion land 
evaporation is more important than has usually been considered. 
Moisture does not take a single flight inland from the water areas, 
but is precipitated and re-evaporated in a series of short flights. The 
forests, utilizing the water in their own growth, again evaporate the 
larger portion of it, acting as a temporary deposit bank from which 



56 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



the moisture may be drawn into the air and redistributed for the 
use of plants in the path of this moisture-laden atmosphere. His 
argument has three steps: First, the coincidence of precipitation in 
eastern United States and the prevailing southerly winds. The cause 
of the direction of the winds has already been indicated. Second, 
evaporation from the land constitutes seven-ninths of the total pre- 
cipitation, while evaporation from the oceans is only two-ninths. 
Third, a cover of vegetation offers a better evaporating surface than 
a body of water and a forest evaporates more water than any less 
dense vegetative cover. There is abundant proof for each of these 
premises and one is compelled to accept the conclusion that the 
forests are an important factor in determining the humidity of the 
prairies. 

To say that there is more abundant moisture in the prairies than 
in the plains is only another way of saying that there is more abun- 
dant vegetation. Many plant geographers, including Pound and 
Clements (1898), C. E. Bessey (1897), Shimek ( 191 1 ) , Gleason 
(1912), Vestal (1914), and others, indicate or actually state that the 
forest is migrating across the prairie. If, then, the eastern prairie 
region is climatically a potential deciduous forest, it appears that the 
eastern forests in their spread can become moisture bearers and there- 
fore equalizers of climatic conditions farther and farther west. As 
the lands growing cultivated crops are, next to the forests, the most 
efficient evaporators, the intensive use of the level lands and the care- 
ful ton-station of the parts not adapted to cultivation in these eastern 
States may make large areas of the country far inland from the 
primary sources of moisture more productive than has hitherto been 
imagined. The examination of the diminishing prairie areas in 
( )hio will doubtless throw light on this important problem. 

( I.I M ATM AM) KdAI'MM - I 7 A(TORS. 

I I e natural Wicce oons of plant associations in a given region have 
been recognized for a lung time. ( owles (igoi) was the first to 
-how thai plan! accessions may be correlated with the physiographic 
ptnenl of a 1" ality. Soil structure, the water-holding capacity 
of the -ml. and the slope of the [and arc determined by historic and 
present physiographic changes. 

Although we now know thai oilier physical factors independent of 
the physiography are steps in the plant succession, nevertheless full 
Credit mUSt be given to ( owle- for presenting so stimulating a view 



waller: crop centers of the united states. 



57 



as the physiographic one for the first systematized studies on plant 
successions. 

In every stage of their development plants respond to the moisture 
and temperature changes of the habitat. The nature of the soil has 
such a far-reaching influence upon plant life that it must be con- 
sidered second in importance to but one factor, the climate. Those 
plant growth factors related to the soil have been named by Schimper 
(1903) the edaphic factors. 

Warming (1909), impressed with the fundamental relation between 
plant growth and available water supply of the habitat, grouped 
vegetation into three principal classes, hydrophytes, mesophytes, and 
xerophytes. The water content of soils was made the basis of his 
work, but when he recently reclassified the three types in order to 
accommodate them more closely to plant distribution, the new system 
was too involved to receive general recognition from plant geog- 
raphers. Schimper made practically the same grouping that Warm- 
ing made of water-content associations. He also pointed out that 
the terms forest, grassland, and desert are a subconscious classifica- 
tion of the principal climatic formations and are only another way 
of expressing the water content of soils. 

The effect of the edaphic factors is to modify the climatic influ- 
ences. The physical and chemical properties of soils tend to diminish 
or intensify the effect of climatic factors upon plant growth. Thus 
we might see in regions of moist climate rock faces, cliffs, or sand 
dunes in which desert conditions would be approached. Xerophytes, 
those plants physiologically adjusted to drought conditions, would 
be able to occupy and hold these situations as long as conditions re- 
mained little suited for the growth of plants requiring more moisture. 

The physical nature of soil structure is more important to plant 
life than the chemical composition of the soils, due to the relation 
between soil texture and water content. An illustration of the im- 
portance of the structure of the soil is seen in the well-known observa- 
tion that the plant successions on clay and on sandy soils are the 
same, but in sand require a much longer time to complete the cycles. 
It should be stated that although there are observations and records 
of soil experimentation carried on by careful agriculturists for more 
than one hundred years, yet much less is known about the soils and 
their effects upon plant life than has been learned of the climatic 
factor. 

The influence of glaciation in destroying topography of former 
times and mixing the soils must be pointed out as important to plant 



58 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



successions and to crops. In the regions of severe glaciation there 
is a widespread homogeneity and locally extreme heterogeneity of the 
soil. Plant successions in these regions are quickly identified. 
Primary successions proceed regularly enough until the climax forma- 
tion, the maple-beech or the mixed mesophytic forest, is reached. 
Secondary successions are readily recognized and these also proceed 
toward the climax. In regions where there was but little leveling and 
mixing of the soil as the result of glacial action, pioneer stages more 
difficult to arrange in succession are frequently met. Regions such 
as those where Fernald (1907) recorded observations on the soil 
preferences of certain alpine plants come under this distinction. In 
older, rougher landscapes in the unglaciated southeastern United 
States investigations conducted by Harper (1914-1917) report the 
plant formations restricted to areas of different soil classes. The 
great local diversity of the vegetation led Fernald to classify plants 
upon a supposed preference for soils of different compositions (based 
on an analysis of the rock origins only), and Harper to correlate plant 
distribution with soil types. 

In all probability, however, the edaphic factors interfere with the 
completion of the successional cycles. It is also likely that temporary 
climaxes have been mistaken for permanent responses. If in the 
southeast, for example, pines are found on sandy and deciduous trees 
on clay soils it should not be supposed that there is an inherent prefer- 
ence of pines for sand and oaks for clay. Rather, it is reasonable 
to believe that the high rate of evaporation and the high rate of 
humus oxidation known throughout the entire southeast and intensi- 
fied on sandy soils arc factors tending to make the temporary climax 
of pines persist for a long time. The climax type of vegetation, then, 
would be composed of evergreens and a large percentage of invaders 
from the deciduous center, since the latter can make more efficient use 
of the light and can offer stronger competition for the better soils. 
Fuller f 1914), working in the Lake Michigan sand dunes, has brought 
raphically the importance of vegetation in modifying the habitat 
by the accumulation of humus, lie shows by many charts the 
Change! in the transpiration rales which result. Similar studies in 
the southeastern center, with simultaneous studies on the retardation 
of humus accumulation due to rapid oxidation, would be interesting 
and profitable. 

The difference in the total annual rainfall and the evaporation in 
*' • t« inprairi< and the deciduous renter is not in itself sufficient 
ount for the great difference in the vegetational aspects. But 



waller: crop centers of the united states. 



59 



the better distribution of the moisture throughout the year produces 
the forests of the East, while the intermittent moisture and long 
drought causes grass to be the dominant type of vegetation of the 
Middle West. The climatic change is gradual from the 6o-percent 
evaporation line eastward. The continuance of the prairie therefore 
is largely dependent upon edaphic conditions. Poor soil drainage 
and the accumulation of muck in the soils seem responsible for the 
black prairie lands. The prairie peninsula (see fig. 5) lies extended 
across Illinois to Indiana because of situations unfavorable to drain- 
age and oxidation. Scattered through the glaciated portions of 
Indiana and Ohio are also small areas of typical prairie. In all of 
these, until drained and broken, big bluestem {Andropogon furcatus), 
one of the bunch grasses typical of the prairies, can be found. 

The plains are more arid than the western part of the prairies and 
much drier than the eastern part. They lie west of the 60-percent 
evaporation line. Short grass, a climax association of Bulbilis- 
Bouteloua, dominates. Just as edaphic prairies are found in the de- 
ciduous forest climax, so are edaphic plains in the prairie climax. 
Gleason (10/10) has demonstrated this in the sand deposits of the 
Illinois River. These deposits date from early post-glacial times 
when the carrying power of the river was severely overtaxed. Blow- 
outs, common in the sand hills of Nebraska, and the same plant suc- 
cessions proceeding from these wind-eroded, bare areas are reported. 

It is interesting to see how climatic and edaphic factors, though 
independent of one another, can sometimes make the same agricul- 
tural practises necessary. An example of this can be noted in the 
custom of furrow planting in the sand plain near Havana, 111., an 
edaphic plains situation, and in western Kansas, the climatic plains. 
In both of these places grains and seeds are dropped into a deep 
furrow in order to be as near to the water table as possible. As the 
plants grow, the hills on either side of the furrow are leveled toward 
it until by the end of the growing season the plants are standing 
in hills. It is also significant to know that the cactus, Opuntia 
rafinesquii, is the common weed of the cornfields in these sand areas. 

The Vegetation Centers. 

From what has been given it can be seen that the forest centers 
are understood to mean those regions in which the combined opera- 
tion of all the climatic and edaphic factors still leaves an environ- 
ment suited to the most favorable development of the species in- 
cluded in the local vegetation type. For example the white pine, Pinus 



6o 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



strobus, the black spruce, Picca mariana, the hemlock, Tsuga cana- 
densis, the balsam fir, Abies balsamca, and the paper-bark birch, 
Bet nla papyrifera, are several trees belonging to the northeastern 
center. A corresponding list of characteristically dominant trees for 
the central deciduous center would include the sugar maple, Acer 
sac char um } the beech, Fagus americana, the chestnut, Castanea den- 
tata, the white oak, Quereus alba, and the tulip tree, Liriodendron 
tulipifera; while for the southeastern center the loblolly, Pinus tacda, 
the long-leaf pine. Pinus palnstris, the sweet bay, Magnolia glauca. 
the bald cypress, Taxodium distichum, and the yellow pine, Pinus 
eehinata. might be chosen. Whether investigators would agree on 
these trees as characteristic of the different centers does not so much 
matter as the point that for each forest center there would be a 
group of trees of different physiological requirements. In the same 
way the lists of the dominant vegetation of the plains, the prairies, 
and the desert would possess marked physiological individuality. 
Dominance in the sparsely vegetated regions does not carry the sense 
of having successfully competed with other plants always, but we can 
speak of dominance because the endemic plants have conquered the 
environment where so many others failed. 

In addition to dominance, Adams ( 1902 and 1909) has pointed out 
maximum size, greatest differentiation of type, and widest range of 
habitat as other criteria of centers of distribution. It is not to be in- 
ferred that the center of distribution is necessarily a place from which 
the species is spreading. Rather it is implied that here the optimum 
climatic factors arc localized and as one leaves a center conditions 
become less than optimum. Therefore those species which are most 
completely dependent upon definite conditions are gradually 
eliminated. 

'I he centers of vegetation are as strongly differentiated by crops 
M by the native forest trees. Timothy, spring wheat, rye, buckwheat, 
•oid DOtatOC OCCUpy the same region as is marked by the first group 
Oi tTC( - in. winter wheat, oats, red clover, and beans dominate 

• 1 n'r.'il region; while cotlon. tobacco, yams, cowpeas, and peanuts 
center in the ioutlu-ast. The same criteria applied above for the 
ton I centei hold for the CfOp centers. The first evidence that we 
m Approaching the center for a given crop is in the number of 
'-ii which it is being grown. The ncxl is that even the rough, 
hilly, and relative!) poor lands produce a fair yield. Then it will 
he noticed that more varietie of tins plant arc known and that indi- 
viduals grow to the greate t size for th.it variety. As one recedes 



waller: crop centers of the united states. 6i 

from a center the converse is true. The varieties, in some cases 
species, become fewer and those most rigidly dependant upon definite 
conditions are eliminated. In even the more tolerant varieties, the 
individuals begin to dwindle in size. Finally, as with the native 
plants the centers of dispersal and the centers of distribution are not 
necessarily the same, so in the crop plants the places where the 
greatest yields per acre are secured and the regions of greatest pro- 
duction may be widely separated. Indeed, in the case of wheat this 
is actually true. 

The fundamental difference in the occurrence of plants belonging 
to the natural vegetation and the crop plants lies in cultivation. 
Man intervenes in behalf of the crop plant which must succeed in 
spite of the unnatural conditions which it faces. This perfectly 
obvious truth is repeated here because of the far-reaching effects of 
plant culture in determining the distribution of the crops when beyond 
their natural centers. As will be shown, this distribution is quite 
contrary to the distribution of the natural vegetation. 

The indigenous plants beyond their centers can be found in the 
poorest habitats only where the struggle for existence is equally 
keen for the invaders from another center and for the members of 
the center. In Ohio, the white pine from the northern evergreen 
center and the scrub pine, Pinus z'irginiana, from the southern ever- 
green center meet on the rock cliffs east of the glacial boundary. 
Neither one of these invaders can compete with the deciduous forest 
species that make up the culminating associations in the better habitats. 
Another example of invaders joining one another in poor habitats is 
the unique association of the cactus, Opuntia rafinesquii, from the 
southwest and the Jack pine, Pinus banksiana, from Canada. The 
two are found together on the sand dunes near Chicago. 

When the crop plants, on the other hand, are to be taken beyond 
their respective centers they must be given the richest land of the 
farms. In Xew York and eastern Pennsylvania when it is desired to 
grow corn the best fields are employed and these have to be further 
reinforced and amended by the use of manures. In the Connecticut 
River valley where tobacco is grown is seen a still more striking 
instance of man's interposition. Not only are the soil conditions 
altered but in order to obtain profitable yields the climatic conditions 
must be altered also by growing the plants under canvas. This 
shades the plants, lowering the transpiration rate during the day, 
while at night it prevents too great a loss of heat by radiation. The 



62 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



climate thus artificially synthesized is more like that which obtains 
in the center of tobacco production. 

The forest centers are plastic groups of plant associations which 
increase or decrease in size, migrating with climatic changes and 
movements of the earth's crust. The crop centers are also motile 
and even less stable than the forest centers. In addition to the 
physiographic and other physical factors which cause the forest 
centers to migrate there are economic developments and shif tings in 
the centers of population which may cause the crop centers to move. 
The problem becomes still further complicated in the case of the 
crops grown primarily for human food. Wheat is an example of 
this. Production is frequently attempted in regions to which the 
plants are not suited. In time these crops tend to become grouped 
about their respective centers. The process of change, however, may 
be a long, slow, and extravagant one. 

One often hears that the great cereal crops are moving westward. 
This is nothing more than the gradual grouping of the centers of 
production of these crops in the area included between the 40-percent 
and 100-percent rainfall-evaporation ratio lines. This is where the 
crops can best be grown. When the United States was first settled 
by the white man he grew the grain he wanted in regions where it 
can not be profitably grown now because of competition with other 
parts of the country where grain production is cheaper than it is along 
the Atlantic Coast. If the census data from 1849 to date are ex- 
amined a steady westward migration of the center of wheat produc- 
tion will be plainly seen. It lias taken considerably more than half 
a century to move the center of wheat production about 100 miles 
north and about 700 miles west or, in other words, from the north- 
eastern evergreen center to the prairie-plains center. When the 
climatic limit ~ fnr all varieties of wheat have been reached the western 
advance, already slowed down, will halt because of limited moisture. 
It is a- yet too earl) t<» predict the effeel in pulling the wheat center 
westward of the intermountain edaphic plains regions. 

'Ivso economic factors which aided the movement of the wheat 
Cntei nblthwetl Where it is now are, first, the invention in 1870 of 
a milling device known as the purifier. By its use a handsomer 
though not a more nutritions Hour could he made from spring wheat. 
Hard ipring Wheat at once jumped from the least desirable class 
of wheat for flour to the most desirable class. The movement of 

■ • towards its center already well begun by the Civil War took a 
definite step northwest as the result of this invention. The second 



waller: crop centers of the united states. 63 

factor is the coming of the railroads into the " World's Breadbasket." 

But the processes of movement have been too haphazard, un- 
certain. Unencouraged by the Federal Government and without 
direction from the experiment stations the movement of crops toward 
their centers goes on too slowly. In the reorganizations sure to 
occur within the next few years it is greatly to be desired that the 
extravagant practices connected with trying to coax a crop from 
plants not growing in the proper habitats be given attention. Crop 
ecology, a not yet developed point of view, will have to be made the 
basis of more intensive studies in crop adaptation and improvement 
if we are to cope more successfully with the food-supply problems 
of the coming generations. Seemingly very few of the investigators 
of crops problems are aware of the importance of the ecological 
researches that have already been and are being carried on. Many 
of these have a direct bearing on crops. 

As soon as we know more about the possibilities of producing a 
given crop in any region the difficulties of marketing to the great- 
est advantage of both the producer and the consumer will be in a 
position to be solved. Labor and transportation adjustments can be 
made when the amount of work to be done and the amount of 
material to be transported can be estimated. 

The Centers of Crop Production. 

We are now in a position to examine the geography of the crop 
centers to discover the relation that they bear to the climatic and 
edaphic factors of the regions in which they are to be found. Be- 
cause of the limited space it has been necessary to pick out a few 
crops only from the many plants cultivated in the United States. 
It will be noticed in those selected that the choice has been on an 
economic basis purely, and not because these plants showed remark- 
ably close agreement with any particular conceptions or theories. 
It would be difficult to substitute readily something else in the place 
of all the crops given. They therefore have distinct economic value. 
If, then, there is correlation with the known vegetation centers, it is 
not because of a prejudice or through skillfully selecting a few plants 
that would be sure to fit. The selection with respect to the biotic 
centers has been haphazard. We would be safe in supposing that 
other crops would show agreement at least as close as those 
presented. 



64 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



A. THE ATLANTIC COAST TO THE GREAT PLAINS. 

The corn and wheat belts agree with the deciduous forest and the 
prairie centers in the United States. Production of these crops is 
greatest between the 6o-percent evaporation line on the west and the 
ioo-percent and no-percent lines on the east (see fig. 5). 

Three sets of factors are operating in combination to establish 
this region as the center for the production of our great cereals. 
These factors may be grouped as climatic, edaphic, and economic. 
While it is impossible to separate and distinguish these groups of 
forces in their actual operation, as canceling any one of them would 
not only destroy the end result but would seriously disturb the 
equilibrium of the other two, yet there is a strong temptation to 
analyze the relative weight of these sets of factors. 

W ithout adding any more to the already lengthy discussions of the 
origin of maize, it is enough to say that it probably had its beginnings 
in the tropics or the subtropics, showed early a remarkable muta- 
bility and adaptability under cultivation so that within perhaps 2,000 
years after cultivation began it reached a wide variety of forms and 
attained a wide geographic distribution. By reason of the hot, al- 
most tropical summers with the relative humidity rather high and 
the annual rainfall sufficient for the growth of the plant, the entire 
area from Ohio to central Nebraska on the north and southward to 
the Gulf of Mexico is suited to corn production. Why, then, when 
the map (fig. 7) is examined do we find the greatest amount of 
production in the northern tier of States? This distribution seems 
<)(\<\ and could only have been arrived at after a good deal of 
trouble in finding varieties which could grow under conditions so 
lar Horn the original environment of the plant. The seven adjacent 
States from Ohio to Nebraska constitute our "corn belt" and pro- 
:X percent of the total quantity grown in the United States. 
Another question which the map provokes is why eastern Illinois 
i> tin- region of greatest production within the center. 

Neither of these questions can be answered simply and directly. 
On the other hand, to go into all the causes and effects would carry 

thi paper far beyond its original Bcope, namely, to show where the 

renter^ of crop production are located and to suggest as briefly as 
' a few oi tin more obvious reasons which brought them 
Where tbejf are. The climatic factor in Indiana and Ohio is suited 
to the profitable production of corn, but production centers in 
fllinoil fof edaphic reasons. The oil in the eastern portion of this 



waller: crop centers of the united states. 



65 




66 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



State is deeper and richer in humus that can be found elsewhere. 
Farther southward, where competition with the cotton crop begins, 
economic reasons prevent a center of corn production from develop- 
ing. The occurrence of cotton south of Kentucky is a competing 
factor for both land and labor. 

In the United States the average annual yield per acre of wheat 
for the period 1 903-1912 is 14. 1 bushels. During that same decade 
England's acre yield seems to have been 32 bushels, Germany's 30.1, 
France's 21, and Russia's but 9.7. There is an almost exact in- 
verse proportion between production and yield per acre that offers 
a fascinating puzzle, for, of the countries named above, Russia and 
the United States are the greatest wheat producers of the nations. 
France, Germany, and England follow, not in immediate, consecutive 
positions but in the order named. 

Wheat has come into universal interest and finds its way into 
every country of the world, which is for it an open market, because 
it has the capability of adapting itself readily to cultivation under 
widely different conditions. In the United States wheat production 
centers on the 60-percent rainfall-evaporation ratio line as can be 
seen by comparing figures 8 and 9 with figure 5. This means that 
the center of wheat production lies west of the best corn lands, 
although on many farms throughout the prairie and deciduous forest 
climaxes both wheat and corn are usually grown if rotations are 
practiced. In the matter of growing wheat in regions too dry for 
com the United States is not an exception to the rule. The great 
wheat-producing regions all over the world are level plains with a 
cool, rather dry climate. It is known that wheat, particularly winter 
wheat, yields larger crops in the more humid sections, yet in normal 
times other crops can be grown in the humid parts of the United 
with greater profit than wheat. It is competition with these 
crops that drives wheat to the plains. 

A dired effect of climate can be seen in the quality of wheat. 
Wheat grown in the cooler, drier climates is, in general, harder and 
darker in color than thai grown in the moister, warmer parts of the 
country. The relative amounts of gluten and starch in the endo- 
perm, determined by the length of the favorable ripening season, 
arc climati responses. KaM and south of the 80-percent evapo- 
ration hue wheat is soft and starchy, with large grains of red, amber, 
or white color. From the 80-percent limits on the east, westward 
' ' 1 I ' '' 'lu re would be an Xo percent evaporation line extended 
if there were no prairie peninsula, the wheat is semihard. On and 



68 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



near the 6o-percent evaporation line the wheat is darker in color 
and harder. It makes a flour that is known as " strong." This 
means that as its gluten percentage is higher, the volume of the loaf 
is greater than from an equal amount of yeast and flour made from 
a " weaker " or starchier kind of wheat. 

An inspection of the spring wheat chart (fig. 9) shows production 
to center on the northern extension of the 60-percent line where all 
the evaporation lines are rather close together and nearly parallel. 
Ecologically, spring wheat could as well be grouped with the crops 
of the northeastern center, but geographically it belongs with the 
prairie climax. Edaphic considerations, then, rather than climatic, 
locate the area of spring wheat production. In Michigan and in Wis- 
consin the climate is as well suited to producing spring wheat as is 
the climate of those States farther west where production centers. 
Where spring wheat and barley are grown we find a great level 
tract of rich soil, a bequest of the old glacial Lake Agassiz. If 
wheat were a native plant indigenous to this climatic section of the 
United States, it would be found here a larger plant and in a greater 
variety of forms than in Michigan, where we may imagine its occur- 
rence also. As however, it is a cultivated crop, the migration of 
which is controlled by man, we can see one valid explanation from 
among many others why its occurrence is limited in the way the 
chart shows. It is again a matter of profits. 

While the eastward distribution is cut off sharply because of 
edaphic conditions, we can see in the climate only a vague and rather 
gen< ral determiner of the distribution north and south. This can be 
interpreted as meaning that physiological races have by no means 
approached the limits of their adaptability and convertability. Spring 
wheat, mostly durum, is found in both Nebraska and Kansas. 

(, at- center slightly north of the corn belt. Climatically, the 
center of production would be expected much farther northward. 
Edaphi rea ons, and the convenience of a spring-sown crop rather 
than a fall sown one to follow corn in the rotation now largely in 
practice in the com belt push the center somewhat to the south. 

Thil southward advance can only he accomplished by the intro- 
duction of small, early-maturing varieties which arc able to make 
■" <nt 11 c of the available moisture. The commonly culti- 
vated Oatl BUppO ed to have ari eil from the Avcna fatua group of 
wild oat , lUCCeed in coo] moist l limatea similar to that of their origin. 
While the origin of smne of tlx- early maturing varieties is at pre- 
Sent Unknown, there is reason to believe that one at least, Hurt, 



JO JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

may have come from Arena barbata. This is a group that offers 
opportunity for valuable genetic and systematic study. 

The region south and east of the 100-percent rainfall-evaporation 
ratio line is ecologically known as the southeastern evergreen center. 
While the rainfall throughout this part of the country is greater 
than it is northward, higher temperatures cause much more rapid 
evaporation. The physiological water requirement is higher. 

Cotton is the principal crop plant of this region. Eastward the 
extension of the southern Appalachians makes too rough a topogra- 
phy for the production of a cultivated crop. Temperature is the 
limiting factor of production northward; moisture is the limiting 
factor westward beyond central Texas. For present purposes the 
southern boundaries of Kentucky and Virginia may be considered 
the limit of cotton production, although there is a slight acreage in 
both of these States (fig. 10). 

There is not space to give in detail all the crops produced in this 
region or to dwell on their ecological significance. The study of the 
maps (figs. 5 and 10) shows beyond doubt that the cotton belt and 
the southeastern evergreen forest are two names for the same region. 
The same influences operating to make this country distinct biologi- 
cally operate in the determination of the crops produced here. For 
other crop charts the reader is referred to the 191 5 Yearbook of the 
United States Department of - Agriculture, from which the maps 
reproduced here were obtained. It is suggested that the same prin- 
ciples of grouping the maps employed here be followed when the 
other crop charts are examined. 

It is interesting to note that although cotton is a cash crop and can 
be converted into money more easily than most crops of the country, 
yet even with this advantage competition with other crops prevents 
cotton from reaching the limit of production. These crops are: 
Tobacco in North Carolina and Tennessee, sugar cane and rice in 
Louisiana, rice in Tcxa* and sweet potatoes, cowpeas, and peanuts in 
of the States included by the cotton belt. 

The northeastern evergreen Forest lies to the north and east of 

tii'- [ IO-percenl evaporation line. It is in area the most extensive 

of the centers, spreading southward down the Appalachians into 

''''•;ia and Alabama, westward through Ontario, northern Mich- 
igan. Wilton in. and Minnesota, and continuing northwest to the 
Bering Strait, with a more northerly distribution than the western 
plant formations. C See fig. 5.) 

I roni the point of view of farm crops this center is restricted to 



72 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



New England and New York, extending through certain sections of 
Pennsylvania. This is the tame hay and pasture region of the United 
States. When it is again noted how closely the prairie-plains climax 
and the northeastern evergreen climax approach one another in the 
Dakotas and Minnesota, the problem of placing spring wheat will 
be appreciated. The reasons for discussing this crop under the 
prairie-plains climax have already been given. 

In Xew York and New England over 50 percent of the improved 
land is in hay or some forage crop (fig. 11), while if the pasture 
land is added to this it will be found that from 80 to 90 percent of 
the improved land is in pasture, hay and forage. 




1 •)'.. 11. Acreage of miscellaneous tame grasses. These include orchard 
redtop, Bermuda, millets, and others. From the 1915 Yearbook of the 
U. S. Dept. of Agriculture. 



The climati< reason for the threat and increasing production of 
fodder in this region is that the lower temperatures make cereal 
production les«- profitable than in the respective centers cf these 
Crop . This brings out clearly the rather strange fact that although 
both wheat and oats are climatically adapted for this center, maize 
is tin- premier cereal of America and seems to serve as a foundation 
upon which the production of the other cereals is built. Edaphi- 
rally tin- thin, -tiff -oik of this center and the uneven topography 
limit the production of crops that must be cultivated. Timothy is 



waller: crop centers of the united states. 



73 



the leading hay crop. Timothy and red clover are grown together 
for mixed hay, but clover alone is not so important in this center 
as it is in the alluvial soils of the lower Ohio and Wabash basins. 

Rye is an important grain crop of the northeastern center and iti 
time of want could be made an appreciable source of breadstuff. 
Since rye bread is already known favorably to many people the 
fancied hardship of having bread that is not snow white could pro- 
bably be overcome with rye flour sooner than with some other kind 
of wheat substitute. In Germany and Russia rye and wheat have 
been used interchangeably for years. Buckwheat also is important in 
this region and if demand came for it, production could be increased. 
It should be noted that buckwheat is the only crop cultivated for its 
edible grain that remained centering in the east during the time that 
the cereals have been carried westward. 

The center of white potato production, w T hile of course being dis- 
persed in the neighborhoods of towns and cities for economic 
reasons that seem to take on more weight than the fundamental 
climatic and edaphic considerations, appears nevertheless to establish 
a fairly close relation with the northeastern evergreen center. This 
means that of the crops now used for that purpose the potato is the 
principal one for human food that is produced in this center. 
Potatoes seem, more than most plants, dependent upon soil conditions. 
Extensive investigations are now in progress to discover the varieties 
best suited to particular edaphic situations. Besides New England, 
another center of potato production is seen in the intermountain 
basin in Colorado wherever the natural moisture or irrigation makes 
its growth possible. 

B. THE PLAINS. 

Turning now to the western half of the United States we find 
that throughout by far the greatest portion of this area evaporation 
consistently exceeds precipitation two or more times. In order to 
live under such drought conditions plants must conserve water in 
an extreme degree. The effect of the limited moisture in determining 
the plant forms which best succeed in the face of this aridity is 
illustrated by contrasting the barrel cactus and the elm. The former 
is almost spherical and compacted into the least possible evaporating 
surface, the latter with its deliquescing trunk melts into many 
branches and leaves spread to the lightest breeze. The cultivated 
plants of the plains climax must be grown under the best known 
methods for saving and utilizing all the water that can be captured 
by the soil and under irrigation. 



74 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Climatically the plains mark a step toward greater drought than 
the prairies, just in the same way that the prairies are more arid than 
the forests. The plains lie west of the 6o-percent rainfall-evapora- 
tion line, extending to the 20-percent line. The latter follows in 
general the north and south trend of the mountains at their eastern 
base (fig. 5). The severe continental climate that prevails is char- 
acterized by the high winds, sudden changes in temperature, and the 
unequal and slight distribution of the precipitation both summer and 
winter. It is this unequal distribution that renders ineffective much 
of the moisture that is precipitated. In the southwestern part of the 
plains where temperature and evaporation are greatest there is a 
gradual and imperceptible merging into desert conditions. 

The direction of the prevailing winds is not such as to bring the 
moisture (in the form of an evaporation-reducing blanket of air, 
rather than actual precipitation) from the Gulf of Mexico, the 
Atlantic, and the forest of eastern America near the Plains. This 
might be stated in another way by saying that after the moisture from 
the east reaches the prairies it is dissipated into the higher air cur- 
rents caused by the absorption and radiation of heat by the great land 
mass of North America. The discussion of " climatic origins " has 
already indicated this feature of continental climates. 

Kansas is divided by the 60-percent evaporation line, so the agri- 
culture of the east and west section of this State may be examined 
for the effects of moisture. In the east there is still sufficient moisture 
for the production of corn and the other crops commonly grown 
under the general methods of farming of the eastern United States. 
West of the line there is not sufficient moisture for this sort of farm- 
ing and a quarter of a century ago the land had no agricultural value 
pt for grazing. It has been pointed out that in Illinois is an 
edaphic plains area that is in agricultural essentials similar to the 
climatic plains region. 

A re\ alu.it ion of the land and a reorganization of agricultural prac- 
ti - ■ ' effected b) tlx- introduction into western Kansas of such 
efficient users Of water as alfalfa, milo, and kafir. Land which was 
regarded 25 year ago as nearly useless cannot be bought under $50 
'in • " now. 'Ilu s> tern of fanning thai must be practiced in these 
lands of little water precludes the possibilities of a crop of even the 
►n erving plants every year, The land is sometimes fallowed 
and allowed to accumulate moisture for a full crop every second 
\ear, as tin li;! been found a more profitable method than growing 
a partial crop each year. 



waller: crop centers of the united states. 



75 




76 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



In western Kansas the soil is permanently dry below a depth of 
5 or 6 feet. In eastern Colorado a depth of 2 feet or perhaps less 
would bring one to the same condition. This comparison not only 
emphasizes the change toward greater aridity but indicates also the 
differences in possible crop production. It must be borne in mind 
that in all of these divisions of the United States there are many 
factors in operation which modify locally the influence of the climate. 
The data from agricultural experiment stations will continue for a 
number of years to be so meager that it will indicate crop possibili- 
ties in only the barest outline. It is safe to predict that production 
is likely to increase in this part of the country. The introduction 
of new plants suited to the conditions is going on at a surprising rate. 
.Methods of studying the environment are improving and cooperation 
between the workers in different fields of endeavor is becoming 
greater. 

Grazing and pasturing and the production of forage are the prin- 
cpial agricultural activities of the Plains climax. The wild hay and 
native pasture grasses must still be given the leading position in crop 
rank (fig. 12). Among the planted crops, alfalfa is easily first. The 
chart of alfalfa production shows its importance near the 60-percent 
evaporation line. Farther west, centers of production can be re- 
garded as irrigated localities, or as regions where the intermountain 
rain and snow make its existence possible. South of the principal 
alfalfa center, but still near the 60-percent evaporation line, coarse 
forage is extensively produced. This part of the southwest plains has 
not reached the semidesert conditions of New Mexico and Arizona. 

Charts of plains crops, i. e., alfalfa, plains wild grasses, and such 
coarse forage as milo and kafir, indicate subdivisions of a region 
..lure the gross physiognomy of the vegetation is largely the same. 
1 he divisions of the plains are ecologically related to the prairie and 
the forest centers in a way that corresponds to the relations brought 
out by the plant geographers studying the prairie-plains vegetation. 

C. THE PACIFIC COAST. 

The third di tind climatic division of the United States is the 

trip -.f trillion from ilie Pacific Coast to the mountains. 
Jt may be rather roughly divided into B northern and a southern 
half. '11"- northern half ifi the region of greatest rainfall in the 
Tinted States more than a hundred inches being recorded as the 
annual average. Cnder the heading of the Pacific Coast the inter- 
mountain basins east of that region may he mentioned because of 



waller: crop centers of the united states. 



77 



their geographic relations with this area. It should be understood 
that in these basins there are great variations in both moisture and 
temperature conditions. In general, it may be said that evaporation 
increases from north to south. This brings out the fact that these 
iritermountain basins have their climatic counterparts in one of the 
three general climatic divisions. When more complete evaporation 
studies have been made, data will be at hand by which the exact 
climatic nature of each of these localities can be determined. 

The proximity to a large water reservoir and the direction of the 
mountain chains and of the prevailing winds are the features which 
express themselves in the abundant rainfall of the Pacific Coast. In 
the northern part where the rainfall is greatest and evaporation least 
a super-forest develops. Douglas fir and one of the cedars form this 
giant forest. Weaver (1914) has traced the development of the 
vegetation in eastern Washington and Idaho and finds that the sub- 
climax is composed of Pseudotsuga and Larix, often with Abies 
grandis. The real climax consists of the cedar, Thuja plicata, con- 
sociation. On the lower levels a scrub forest, or as it is more fami- 
liarly known, the chaparral, predominates. We are safe in believ- 
ing that in the scrub forests evaporation begins to exceed rainfall and 
the moisture obtainable from the melting snows. In the extreme 
south and not a great distance from the coast is the desert. 

The topography limits the distribution of the sequoias, which are to 
be found only near the mouths of the canons where there is exposure 
to the foggy atmosphere, protection from excessive evaporation, and 
the possibility of obtaining water circulating underground within 
reach of their roots. Topography also determines largely the 
evaporation and so controls the vegetation of the intermountain 
valleys. Some places are semideserts, even approaching desert con- 
ditions, e. g., the Snake River. Others present edaphic plains, prairie, 
or forest climaxes. 

The northern part of the Pacific Coast is primarily a hay-prdoucing 
region just as is New England. There is a notable feature peculiar 
to this region due to the abundant rainfall. It is the customary prac- 
tice to use for hay plants the same crops which in the Middle West 
are grown mainly for their grain. Wheat and barley are used as 
forage crops and are classed by the census enumerators as "grains 
cut green." It is also possible that some day the center of the bulb- 
growing industry for the United States may be located in favorable 
situations in this part of the country. The apple industries of the 
moister river valleys of Washington, Oregon, and Idaho are famous 



78 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



all over the world. In a number of localities in this region enough 
wheat is grown to make an important wheat center. 

In the southern part of the Pacific Coast region annual crops are 
largely replaced by tree crops, walnuts, citrus fruits, olives, and 
other perennials. 

Animal Centers. 

This account would not be complete without some reference, how- 
ever slight, to the animal industries that follow the production of 
plants. The reader is urged to examine the charts on livestock in 
the 1915 Yearbook of the United States Department of Agriculture 
and compare them with the rainfall-evaporation ratio chart and the 
crop centers. 

Adams (191 5) has shown that the insect communities of a forest 
and a prairie are totally different. This indicates the same adjust- 
ment to physiological requirements that has been seen in the plants. 
It also indicates something of the dependence of animals upon certain 
plants for their food. The interrelations between plants and animals 
growing out of this fundamental dependence are enormous. Adams's 
report will have to be read in full to appreciate something of it. 
Back of the interrelations between plants and animals is the rela- 
tion of both to the physical factors of their environment. 

The presence of certain wild plant species in a locality is suffi- 
cient to account for the occurrence of insect species (Adams, 1914, 
loc. cit., p. 46). It is a more widely known fact that the introduction 
of wheat into America was shortly followed by the introduction of 
the Hessian fly. This insect has spread everywhere by following 
the path of its favorite food plant. It is almost a foregone conclu- 
sion thai a map showing the limits of wheat distribution would also 
-how the limits of the distribution of the Hessian fly. The Colorado 
potato hectic was able to spread to all parts of the United States 
' pathf had been made for it by planting potatoes and one of 
these happened to touch a natural center for the beetle. Such in- 
stances as these could easily be multiplied. 

Wt might perhaps be inclined to believe that the adjustments of 
a nim a l l to plant development could most easily be found among the 
in* t . a numerous and highly specialized group. I)iit the responses 
ivei al and are to be seen whether we go up or down the evolu- 
tional-; - ale. Adams reports prairie and forest spiders and snails, 

while Hanldnsoi) ( 1915) distinguishes the forest vertebrates, rang- 
ing from fish to mammals, from the prairie vertebrates, ranging from 



waller: crop centers of the united states. 



79 



amphibians and reptiles to mammals. The work of Thompson- 
Seton (1909) on the North American mammals should also be con- 
sulted in this connection. Many charts of both herbivorous and 
carnivorous animals, together with many notes on their distribution, 
are presented in the two volumes of his work. These maps show 
that the biological centers may be examined from many angles, de- 
pending upon the particular field of endeavor of the worker, but that, 
after all, the centers are expressions of the same interactions of 
climatic and edaphic factors. The distribution of the striped ground 
squirrel, Citellus tridecemlineahis, marks the prairies. Its range 
carries it across Illinois eastward into Indiana and Ohio, and it is also 
seen in the sand plain of southern Michigan. In the plains, the prong- 
buck, Antilocapra americana, has its food habits fixed by its environ- 
ment. In captivity attempts to give it other food than the familiar 
grass, cactus, and sagebrush have proven unsuccessful. 

Much more readily observable is the dependence of the domesti- 
cated animals upon the cultivated plants. Reference to the charts 
in the 1915 Yearbook will show that the dairy industries are located 
in the northeastern evergreen center and on the Pacific Coast. These 
regions it will be remembered are the natural tame hay and pasture 
centers of the United States. Economic reasons also enter into this, 
of course, but the significance of a cool moist climate with an abun- 
dant production of forage reasonably certain every year is too large 
a fundamental fact to overlook. Beef cattle and swine are found 
centering in and slightly west of the corn belt. Their relations to the 
great grain-growing areas are not difficult to perceive. The greatest 
production of horses is in the region just north of the corn belt. This 
is the present center of oats production also. Mules are supposedly 
sturdier work animals than horses. They are found centering in 
the cotton belt because, more in the past than at present, grain and 
fodder was not produced in the cotton belt and so good feed was 
difficult to get. Also, as their drivers were less likely to be careful 
of work stock, there was a greater chance of survival after ill 
treatment. 

Sheep are abundant in the arid regions. Though production is 
greatest in the West, there is also an important center in Ohio and 
Pennsylvania. On the whole this distribution is most interesting and 
brings out several important facts. First it should be noted that the 
sheep of the East and Middle West are more likely to be mutton 
sheep and the wool produced is only a byproduct. Unfortunately 



8o 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



there is no way of determining easily just how far this is true. The 
distribution in the West must be regarded as climatic, since sheep 
can find a living from vegetation on which other animals would 
perish, and the western distribution shows sheep to be limited to the 
rather sparsely vegetated areas. Further, not only does climate de- 
termine the distribution of sheep, but it also controls the method of 
handling them. Because there is not enough vegetation to feed 
them in any one place, they must be driven to follow the growth of 
vegetation which springs up after the rainy seasons. This is espe- 
cially true in Wyoming and Montana, the two greatest wool-producing 
States, and in the drier portions of Oregon and southern California. 

The center in Ohio and Pennsylvania is significantly located east of 
the glacial boundary, where the rougher topography favors pasturing. 
In addition to pasture, access to grain and fodder is not difficult and 
the feeding of the sheep in winter is a part of the method of handling. 
The eastern Ohio and western Pennsylvania center of sheep raising 
is essentially an edaphic center. 

Many of the statements in the foregoing paper can not be appre- 
ciated fully until the relation of the crop centers to the centers of the 
natural vegetation has been completely analyzed. The intensive 
methods used by the ecologists in the study of the habitat, namely, the 
use of instruments for the exact observation of the moisture and tem- 
perature and photographs to record plant growth in relation to the 
surroundings, can not be too strongly emphasized. In a number of 
instances the measurements found by ecologists who have used 
standardized vaporimeters are applicable to agricultural studies. In 
all studies of evaporation, edaphic factors sometimes operating with 
and sometimes against the climatic, deserve the most careful in- 
terpretation. These in turn are dependent upon the present and 
historic geology and topography. 

SUMMAKV. 

The crop ( enter- of the United States agree with the biotic centers. 
In detail tin means that the corn and winter wheat belts correspond 
tO tlf deciduoui central forest and the prairie climaxes, the tame 
i ; -i tun region to the northeastern evergreen forest, the cotton 
belt to the southeastern evergreen forest, and so on. The rainfall- 
evaporation ratio map is useful for the demarcation of these centers 
' )" 't ;n'e included four factors of climate, namely, relative 
humidity, temperature of the evaporating surface, and wind velocity 



waller: crop centers of the united states. 8i 

as the divisor, and precipitation as the dividend. These four factors 
are of profound importance to plant growth. 

Edaphic factors frequently determine the distribution of the culti- 
vated plants. Edaphic and climatic factors, although they may be 
independent of one another in their operation, sometimes cause the 
same agricultural practices to be employed. Economic factors 
modify the influence of climate and soils. 

A fundamental difference between crop plants and the natural 
vegetation is seen when plants are found beyond their usual centers. 
The crops are found on the best soils only, since that is their sole 
chance to compete with other crops for profit. Plant invaders of 
the indigenous vegetation migrating from their centers can offer com- 
petition in the poorest habitats only. In the better habitats the plants 
belonging to the center are little influenced by invaders. 

In addition to the exotic crops being given the best fields, further 
soil modifications are usually introduced. In the extreme cases, 
climatic as well as soil modifications are practiced. Field plants are 
then grown on a comparatively large scale under glass or cloth shelter. 

The domesticated animals are grouped about the centers of pro- 
duction of those crops upon which they are most dependent. 

The methods used in studying plant succession have been used here. 
It is in this field of research that an accurate interpretation of condi- 
tions as consequences of the operation of physical forces of the past 
and present has been made. Migration, including invasion and com- 
petition, the latter implying dominance, are the direct results of inter- 
action of climate and soils upon vegetation. 

Literature Cited. 

Adams, C. C. 

1902. Southeastern United States as a center of geographical distribu- 
tion of flora and fauna. In Biol. Bui., 3: 115. Literature cited, 
p. 122. 

1908. An ecological survey of Isle Royale, Lake Superior. In Rpt. 
Mich. Bd. Geol. Survey for 1908. 

1915. An ecological study of prairie and forest invertebrates. Bui. 111. 

State Lab. Nat. Hist'., v. 9, art. 2, 280 p., 43 P 1 -, 18 Litera- 
ture cited, p. 46-47, 66-98, 1 19-122, 157-159- 

Bessey, C. E. 

1897. Are the trees advancing or retreating on the Nebraska plains? 

In Science, 10 : 768. 
1899. The forests and forest trees of Nebraska. In Rpt. Agr. Nebr. 

1900. Literature cited, p. 79. 



82 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Covvles. H. C. 

1901. Physiographic ecology of Chicago and vicinity. In Bot. Gaz., 31 : 
73-108; 145-182. Literature cited, p. 145-150. 

191 1. The causes of vegetative cycles. In Bot. Gaz., 51: 161-183. Lit- 

erature cited, p. 167-182. 
Ferxald. M. L. 

1907. The soil preferences of certain alpine and subalpine plants. In 
Rhodora, 9: 149-193. Literature cited, p. 171-176. 

Fuller, G. D. 

1914. Evaporation and soil moisture in relation to the succession of 

plant associations. In Bot. Gaz., 58 : 192-234; 27 fig. Literature 
cited, p. 232-233. 
Gleasox, H. A. 

1910. Vegetation of the inland sand deposits of Illinois. Bui. 111. State 
Lab. Nat. Hist'., v. 9, art. 3, 174 p., 20 pi., 6 fig. Literature cited, 
P- 77. 

1912. An isolated prairie grove and its phytogeographical significance. 

In Bot. Gaz., 53 : 38-49. Literature cited, p. 45-49. 
Haxkixsox, T. L. 

1915. The vertebrate life of certain prairie and forest regions near 

Charleston, 111. Bui. 111. State Lab. Nat. Hist, v. II, art. 3, p. 
281-303, pi. 64-79. Literature cited, p. 284-301. 

Harper, R. M. 

1914. Geography and vegetation of northern Florida. In 6th Ann. Rpt. 
Fla. State Geol. Survey, p. 163-451, figs. 90. Literature cited, p. 
174, 211, 217, 221, 238, 277, 291, 304, 315. 

Hkxry, A. J. 

1906. Climatology of the United States. U. S. Dept. Agr., Weather 

Bur. Bui. Q, 1012 p., 33 pi. Literature cited, p. 7-84. 
Jonxsox, K. S. 

1907. Mian monthly and annual relative humidity charts of the United 

States. In Rpt. South African Assoc. Adv. Sci., p. 161-168, 
charts 13. 
Livingston, B. E. 

[915, A single climatic index to represent both moisture and tempera- 
ture conditions as related to plants. (Abstract from paper read 
during latest Columbus meeting, Amor. Assoc. Adv. Sci.) 
Poi nd, \<„ and Clements, F. E. 

jKoK. The vegetation regions of the prairie province. In Bot. Gaz., 25: 
381 394, Literature cited, p. 347. 

Russell, T. 

jKKK. Evaporation hi U. S. Dept. Agr., Monthly Weather Review, 
September. 

&AMUIT, C. S. 

1884. Forests f»f North America. /" Rpt 10th Census of the U. S„ v. 9. 
S< him I'm, A. F. \V. 

190$ I Mant geography on a physiological basis. Translated by W. R. 
I'islier. Oxford. Literature cited, p. 3 7. 



albrecht: changes in soil nitrogen content. 



83 



Shimek, B. 

1911. The prairies. Bui. Lab. Xat. Hist. Univ. Iowa, 6: 169. 
Transeau, E. N. 

1905. Forest centers of eastern America. In Amer. Xat., 39: 875-889; 
6 fig. Literature cited, p. 884-886. 

Warming, E. 

1909. Ecology of plants. Oxford. 
Weaver, J. E. 

1914. Evaporation and plant succession in southeastern Washington and 
adjacent Idaho. In Plant World, 17: 273. 

Zon, R. 

1913. The relation of forests of the Atlantic plain to the humidity of 
the Central States and prairie region. In Science, 38 : 63. 



CHANGES IN THE NITROGEN CONTENT OF STORED SOILS. 1 

Wm. A. Aldrecht. 

Certain determinations for soil nitrogen at the Missouri Agricul- 
tural Experiment Station have shown marked evidence of increases 
having taken place during storage. As the soil samples in question 
had been stored in containers which were not airtight and as some 
of the samples were moist, the possibility of bacterial fixation at first 
suggested itself. As the room where the samples had been stored 
adjoined the laboratory in which more or less ammonia was being 
used it seemed more probable that the increase was due to direct 
ammonia adsorption. 

Literature on the adsorption of gases by soils puts ammonia as one 
of the gases most easily adsorbed. Schlosing found that a moist 
soil exposed to the air adsorbed nitrogen at the rate of 38 pounds 
per acre per annum, which was mainly in the form of ammonia. 
This adsorption took place whether the soils were acid or alkaline, 
dry or wet (7). 2 A. D. Hall, of Rothamsted (1), after studying 
ammonia absorption from the air by means of sulfuric acid, states 
that " the maximum absorption per annum amounts to less than a 
pound per acre." In testing the amounts of ammonia that soils will 
adsorb from an unlimited supply, Muntz and Gerard found that 1 
kilogram of garden soil took up 5.38 grams of ammonia (4). The 

1 Contribution from the Department of Soils, University of Missouri. Re- 
ceived for publication September 4, 1917- 

2 Figures in parentheses refer to papers similarly numbered in "Literature 
cited," page 88. 



8 4 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



presence of various substances in the soil increases the adsorptive 
capacity for ammonia. It has been shown that calcium carbonate 
(7), ferric hydroxide, and organic matter (2) have marked effects, 
while moisture is not as effective as one would suppose (6). 
Lyon, Fippin, and Buckman (3) state that — 

Absorption of gases by soils is largely an adsorption phenomenon, the gases 
being condensed on the surface of the particles. The absorption is greater, 
the finer the particles of soil, but this increase is not directly proportional to 
the increase in surface, large particles apparently having a greater absorptive 
power than their surface would indicate. 

As a surface phenomenon the amount adsorbed will be dependent 
on such factors as partial pressure, temperature, viscosity of the 
vapor, physical condition, chemical composition, and others. When 
the gases are once adsorbed, they are maintained with marked tenacity, 
as is shown by glass, which holds hygroscopic moisture at tempera- 
tures as high as 500 C. Sufficient evidence is available to emphasize 
the fact that the soil is a powerful adsorber of gases. This mass of 
information suggested the possibility of contamination of stored 
samples through this means. The following study was undertaken 
to find out whether contamination by adsorbed ammonia or bacterial 
activity was responsible for variation of the nitrogen in stored 
samples. 

The plan of storage and treatment of the soil was as follows. Two 
soils widely different in nitrogen content were collected ; one a Shelby 
silt loam with 2,325 pounds, and the other a Summit clay loam with 
7,950 pounds of nitrogen in 2,000,000 pounds of surface soil. Each 
sample was thoroughly mixed and divided into three parts. On one 
part the determinations of nitrate, ammonia, and total nitrogen were 
made as soon as possible after collecting. The second part was put 
into a room adjoining the general soils laboratory and spread out on 
a table Determinations of the three forms of nitrogen were made 
four weeks later. The third portion of the soil was put into bags in 
the moi^l condition and stored in a basement room to dry slowly for 
analysis two months later. 

] he analytical methods were those commonly used. Nitrates were 
determined by extracting the over dried soil with [J hydrochloric 

add, making it alkaline, boiling off the ammonia, reducing with 
Devarda's metal, and distilling. The ammonia was measured by 
distilling the -oil and magnesium oxide with compressed air and 
' hile the total nitrogen determination was according to the 



ALBRECHT : CHANGES IN SOIL NITROGEN CONTENT. 



85 



official method modified for nitrates by use of sodium thiosulfate. 
The samples were thoroughly mixed each time and sieved through 
an 80-mesh sieve for total nitrogen determinations and through a 
20-mesh sieve in the other cases. All determinations were calculated 
on a water-free basis. Each sample for total nitrogen was dried in 
an oven at 107 C. for eight hours and then transferred to a flask for 
digestion. Calculations of pounds per acre were based on each 
separate water-free sample. For receiving nitrate and ammonia 
distillations, a ^ sulfuric acid was used and duplicates checked 
within 0.2 c.c. for the nitrates but not so closely for ammonia. For 
total nitrogen analysis a ^ acid was used, and determinations again 
checked to 0.2 c.c. All samples except in a few cases were run in 
quadruplicate. 

Table 1 gives the data from the different nitrogen determinations. 
The figures are averages calculated from four determinations. 



Table i. — Variation in amounts of ammonia, nitrate, and total nitrogen of a 
soil stored under different conditions. 

AMMONIA NITROGEN. 



Date of determination and 
place stored. 


Silt soil. 


Clay soil. 


Weight of 
water-free 
soil. 


Nitrogen. 


Weight of 
water-free 
soil. 


Nitrogen. 


Dried 28 days near labora- 
Dried 56 days in basement. . 


gms. 
82.60 

83.28 

82. 70 


mgs. 
i.53 a (i-48-i.58) 

2.31(2.14-2 50) 
1.45(1.32-1.60) 


gms. 
71-55 

80.37 
78.95 


mgs. 
1.50(1.30-1.73) 

2.66(2.57-2.72) 
1.58(1.47-1.73) 


NITRATE NITROGEN. 


On day sampled 

Dried 28 days near labora- 
tory 

Dried 56 days in basement. . 


61.97 

63.68 
63.24 


Trace 

Trace 
Trace 


53.66 Trace 

61.45 : o.3io(.255-.357) 
60.37 I o.34o(.255-.4o8) 


TOTAL NITROGEN. 


On day sampled 

Dried 28 days near labora- 
tory 

Dried 56 days in basement. . 


9-79 

9.80 
9-73 


10.96(10.69- 
11. 12) 

11.32(11.19- 

11.49) 
11.21(11.09- 

11.38) 


9.480 

6 9-46 
9.29 


37.64(37.45- 
37.84) 

39.32(39.19- 

39-48) 
37.77(37.58- 

37-92) 



a Figures in parentheses denote limits of variations in figures from which 
averages were calculated. Variations in hygroscopic moisture determination 
were less than 6 milligrams for the 10 gram samples. 

6 Three determinations only. 



86 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



The total nitrogen on the day sampled in the silt soil was 2,237 
pounds per acre in 2,000,000 pounds of surface soil, variations 2,183 
to 2.263 pounds; in the clay soil, 7,943 pounds (7,901-7,981). In 
the soils dried 28 days in the laboratory, the total nitrogen was, silt 
soil, 2,311 pounds (2,284-2,345), and clay soil, 8,315 pounds (8,286- 
8,350) . In the soils dried 56 days in the basement, the total nitrogen 
was. silt soil, 2,306 pounds (2,280-2,339), and clay soil, 8,133 pounds 
(8,097-8,164). 

The data given for the ammonia nitrogen in Table 1 indicate an 
increase in this form when the soil was dried near the laboratory, 
while there was no increase when stored in the basement room. This 
indicates contamination by gaseous ammonia rather than bacterial ac- 
tion, for if the latter agent had been responsible the sample stored 
in the basement should have given an increase in ammonia also. 
Bacteria seemed to be playing no role in a measurable way. 

The nitrates remained largely unaffected by storage, even though 
the clay loam shows less nitrogen as nitrate on the day of sampling 
than when stored. No explanation for this is offered. 

The total nitrogen showed an increase with storage, particularly 
w hen the soil was kept near the laboratory. With silt soil this in- 
crease was small, but with the clay loam quite significant, though 
the determinations are a trifle erratic in the latter case. The data as 
a whole indicate that there is no significant change in the nitrogen by 
bacteria. 

I he above results prompted another series of analyses on the silt 
soil only, to test the possibility of contamination of both moist and 
dl •;. wils by ammonia. Several portions of a large sample of soil 
were treated as follows. One part was analyzed for ammonia and 
total nitrogen ,> -non after sampling as possible. A second portion 
nrai dried in the laboratory where ammonia was used. The third 
part was dried in a greenhouse located in an orchard away from any 
ammonia. As soon as the soil iii the greenhouse was well dried, 

analyses were made on ome of it, and the remainder was divided 

into halves, leaving one half in the greenhouse and transferring the 
other halt in the laboratory where it was spread out near an 

evaporating dish containing aboul <><> c.c, of ammonium hydroxide. 

'I hi. procedure Was followed since no ammonia was being used in the 
room ;it thai time. 'I he data are given iii Table 2. Determinations 
were made in w|. of four and figures given are averages. 



albrecht: changes in soil nitrogen content. 



Table 2.— Ammonia and total nitrogen in silt soil as affected by different con- 
ditions of storage. 



Time of sampling and 
place stored. 


Ammonia nitrogen. 


Total nitrogen. 


Water- 
free soil. 


Nitrogen in 
sample. 


Water- 
free soil. 


Nitrogen in 
sample. 


Nitrogen in 
2,000,000 pounds 
of surface soil. 




Grams. 


Mgs. 


Grams. 


Mgs. 


Pounds. 




79-73 


1. 01 


9.6609 


II.248 


6 2,328 






a (o.86-i.i2) 




(11.189- 


(2,317-2,335) 


Dried in greenhouse 10 days . 


77-85 


1.23 


9.7320 


11.288) 
II.275 


2,316 






(1.07-1.37) 




(il. 189- 


(2,298-2,340) 










11.388) 




Dried in laboratory 10 days. 


77-91 


2.91 


9.7382 


II.474 


2,356 






(2.75-3.11) 




(n-437- 


(2,348-2,370) 










H-537) 




Dried and stored in green- 














78-13 


1.72 


9.7662 


11.409 


c 2,337 






(1.65-1.78) 




(n-371- 


(2,328-2,347) 










11.468) 




Dried in greenhouse and 












stored in laboratory near 








• 






78.30 


60.43 


9.7878 


19.041 


3.890 






(60.04-60.81) 




(18.823- 


(3,846-3,906) 










I9-II4) 




Dried in greenhouse and 












stored near ammonia; 












samples not heated 






V7878 


18.823 


c 3.846 










(18.678- 


(3,816-3,876) 










18.969) 





n Numbers in parentheses denote range in variations. Hygroscopic moisture 
determinations varied no more than 6 milligrams in a 10-gram sample. 
h Three determinations only. 
c Two determinations only. 
d Water-free soil figured from sample above. 



The data in Table 2 show 'beyond a doubt that both the moist and 
dry soils have taken up nitrogen as ammonia when stored in a labora- 
tory in which ammonia fumes are present in considerable amounts. 
The sample dried in the greenhouse and then exposed in a dry state 
to ammonia fumes adsorbed enough to give 1,553 pounds per acre 
increase in the total nitrogen. In the distillable ammonia there was 
an increase by drying in the laboratory and an unusual increase for 
the dry soil left near an ammonia container. The increase in am- 
monia' nitrogen over the sample when first collected amounts to 1,517 
pounds per acre, corresponding closely to the increase of 1,553 pounds 
per acre of total nitrogen. Evidently the use of small amounts of 
ammonia in the laboratory is sufficient to increase markedly the 
nitrogen in a soil exposed there. That the ammonia is not held by the 
moisture present in the soil is indicated by the increase of nitrogen 



88 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



when the dry soil was brought from the greenhouse and stored near 
the ammonia. This soil had but 2.75 percent of moisture, yet took 
up the equivalent of 1,517 pounds of nitrogen per acre. Ammonia 
thus taken up is held strongly enough so that heating for eight hours 
at 107-108 C. does not drive it off. This is shown by the figures 
for those samples analyzed without first heating them in an oven as 
compared with those heated. That the nitrogen is taken up in the 
form of ammonia is shown by the fact that the increase of this ele- 
ment is distillable with magnesium oxide. In the form of ammonia, 
the nitrogen is held, not by moisture, but by adsorption as a purely 
physical phenomenon. 

This brief study indicates that in case of soils stored in or near a 
laboratory in other than air-tight containers, there is a grave danger 
of contamination by ammonia, whether the soil is wet or dry. It 
indicates further that when moist soils are left to dry slowly there is 
little danger of bacterial action measurably affecting the nitrogen 
content. 

Literature Cited. 

1. Hall, A. D., and Miller, N. H. J. 

On the absorption of ammonia from the atmosphere. In Jour. Agr. 
Sci., 4: 56-68. 191 1. 

2. HlLGARD, E. W. 

Soils, p. 274-275. 1006. 

3. Lyon, T. L. f Fippin, E. O., and Buckman, H. O. 

Soils, their properties and management, p. 367. The Macmillan Co., New 
York. 1 916. 

4. Muntz and Gerard. 

Cited by A. Hebert. in Expt. Sta. Record, 5: 144. 1893. 

5. Patten, H. E., and Gallagher, F. E. 

Absorption of vapors and gases by soils. U. S. Dept. Agr., Bur. Soils 
Bui. 51, p. 49. 1908. 

h. I\H< MARDT and Bl.OOMTRITT. 

Jour. F. prak, Chemic, 98: 167. Cited by Hilgard, E. W. (2), p. 276. 

7. SCHLCSSING, T. 

( omp. Haul. (Paris) no: 429 490. /lbs. in Expt. Sta. Record. 3: 110- 

112 189I. Cited by King, F. II.. The Soil, p. 122. Also cited by 

Patton and Gallagher (5), p. 30. 



BALL & CLARK '. NAMING WHEAT VARIETIES. 



8 9 



NAMING WHEAT VARIETIES. 1 

Carleton R. Ball and J. Allen Clark. 

Crop varieties must be distinguished by names. These names 
must be used frequently by a host of agronomic workers as well as 
by crop growers and crop users. The form and appropriateness of 
these names, therefore, are of general interest. It is desirable that 
they be short, simple, and appropriate, easily spelled and pronounced. 
It also is desirable that a single name of this kind be designated and 
accepted for each recognized variety. 

Confusion in Varietal Names. 

The multiplication of names and other designations for crop varie- 
ties has been carried to great extremes. The resulting confusion 
also is very great, especially in those crops like wheat where the 
number of actual varieties is very large. ' These names and near 
names may be classified into three series, as follows: (1) Names, 
(2) descriptive phrases, and (3) numbers. As examples of names, 
Fulcaster, Fultz, Jones Fife, and Kubanka may be cited. As examples 
of descriptive phrases we may quote Bluestem, Early Red Clawson, 
Jones Paris Prize, Purple Straw, and White Australian. Numbers 
applied in place of names may be typified by Iowa No. 404, Min- 
nesota No. 163, and Washington Hybrid No. 128. 

At the present time, the existing confusion and multiplication of 
varieties places a great burden on agronomic workers. It renders 
uncertain and difficult the interpretation of published results of ex- 
periments. This confusion occurs in two principal ways. (1) The 
same name is applied to very different varieties in different parts of 
the country; (2) The same variety passes under several different 
names in different parts of the country, or even in the same part. 

Good examples of the same name, or rather descriptive phrase, 
applied to different varieties are Bluestem and Red Russian. In the 
Far West, Bluestem is an awnless variety with glabrous white glumes 
and white soft kernels, usually spring-sown. According to Leighty, 
an eastern fall-sown variety with similar spike and kernel characters 

1 Contribution from the United States Department of Agriculture, Washing- 
ton, D. C. Presented by the senior author at the tenth annual meeting of the 
American Society of Agronomy, Washington, D. C, November 14, 1917. 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



is known as Kentucky Blnestem. In the upper Mississippi Valley, 
Bluestem is a spring-sown awnless wheat w r ith pubescent white glumes 
and red hard kernels. In the eastern United States, the names Ala- 
bama Bluestem and Georgia Bluestem are applied to fall-sown awn- 
less wheats with glabrous white glumes and red, midsized, soft 
kernels, while Pennsylvania Bluestem differs in having brown glumes. 

The descriptive phrase-name, Red Russian, is applied to four and 
perhaps five different varieties in this country. In the humid dis- 
tricts of the Pacific Northwest it is applied to a fall-sown variety 
with large, clavate, awnless spikes, glabrous white glumes, and red 
large soft kernels. In the Great Plains area and westward, the name 
Red Russian is commonly applied to the Russian winter wheats of the 
Crimean group, as Crimean, Kharkov, and Turkey. These have small 
fusiform awned spikes, white glabrous glumes, and red midsized hard 
kernels. In the northern part of the Great Plains area, the name 
Red Russian is applied to a spring-sown variety, similar in general 
appearance to the Crimean winter wheats, but differing in spring 
habit, glume characters, .and semihard kernels. A fourth variety 
called Red Russian is grouped by Leighty with those eastern wheats 
having awnless spikes with glabrous brown glumes and red soft 
kernels. The list of names which have been applied to at least two 
different varieties is too long to present here. 

The second case mentioned was where the same variety passes 
under two or more varietal names. Here the real difficulty is to 
prove that the varieties bearing different names are really identical. 
This is a much more difficult task than determining strikingly evi- 
dent differences, and is completed only after careful study and com- 
parison. Here too we must recognize that there are differences of 
performance not necessarily correlated witli visible characters. For 
such case- the nccssary allowance must be made. 

There remain, however, many cases in which the identity of varic- 
tii bearing two Or more different names is evident. The Crimean 
group " J hard red winter wheats, the dominant crop in Kansas, 
Nebraska, and ( Jklahoma, is a good illustration. Alberta Red, 
can. Kharkov, Malakov, l\cd Russian, Torgova, and Turkey 
are only different names for a single variety. It may be called 
Crimean, Kharkov, or Turkey with e<|ual accuracy. Bcloglina, 
00 the other hand, | an be se] >a 1 a t <•,] From these others on one minor 
character, namely, the longer beaks and s<|iiarer, more deeply notched 
•houl'l'-r- of tlx- flumes. Kanred. a pure line separated from 



BALL & CLARK! NAMING WHEAT VARIETIES. 



91 



Crimean, C. I. No. 1437, was bred and named by the Kansas Agri- 
cultural Experiment Station. It is, however, a true Beloglina, and 
differs from other Beloglinas only in consistently higher yield, so far 
as known. 

Another good illustration is the Pacific Bluestem, the dominant 
spring variety of the Columbia and Snake River basins in the Pacific 
Northwest. Research has shown it to be identical with the White 
Australian, which has been the dominant variety of California for 
fully 60 years. 

In addition to the different kinds of confusion of names which have 
been discussed, there are many examples of objectionable names of 
other sorts. Many varietal designations are long and cumbersome 
descriptive phrases ; for example, Early Red Clawson, Jones Silver 
Sheaf Longberry, etc. Others are equally long and cumbersome 
numbers, as, for example, Minnesota No. 169, Nebraska Hybrid No. 
28, or Washington Hybrid No. 143. Another disadvantage in using 
numbers as names is that an error in a single numeral renders the 
variety unrecognizable. 

The facts and conditions set forth in the preceding discussion can 
be amplified almost without limit. They seem to the writers to show 
the need for some concerted action on the part of agronomic workers. 
The writers are about to begin the publication of a classification of 
wheats. In it varietal names must be used and confusion in their 
use avoided. This means that duplication of the same varietal name 
for different varieties can not be recognized. Conversely, different 
names for the same variety must be eliminated. 

A Proposed Code of Nomenclature. 
A brief but comprehensive code of nomenclature is presented here- 
with for the consideration of the members of the American Society 
of Agronomy. It is hoped that it may be adopted in some form at 
this meeting, so that the authors may have opportunity to select 
varietal names for their classification in accordance with its rules. In 
this way, whatever names are used would have the backing of a 
responsible body of agronomists. 

Code of Nomenclature. 

The following rules governing the naming of varieties of crop 
plants are hereby proposed for consideration and adoption by the 
American Society of Agronomy, at the annual business meeting on 
November 13, 1917. 



92 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



1. Eligibility to Naming. No variety shall be named unless (a) distinctly 
different from existing varieties in one or more recognizable characters, or (b) 
distinctly superior to them in some character or quality, and unless (c) it is to 
be placed in commercial culture 

2. Priority. No two varieties of the same crop plant shall bear the same 
name. The name first published (see Rule 4) for a variety shall be the ac- 
cepted and recognized name, except in cases where it has been applied in viola- 
tion of this code. 

A. The term, " crop plant," as used herein, shall be understood to mean • 
those general classes of crops which are grouped together in common usage 
without regard to their exact botanical relationship, as corn, wheat, sorghum, 
cotton, potato, etc. 

B. The paramount right of the originator, discoverer, or introducer of a 
new variety to name it, within the limitations of this code, shall be recognized. 

C. Where the same varietal name has become thoroughly established for 
two or more varieties, through long usage in agronomic literature, it should 
not be displaced or radically modified for either one, except where a well- 
known synonym can be substituted. Otherwise the varieties bearing the same 
name should be distinguished by adding some suitable term which will insure 
their identity. 

D. Existing American varietal names which conflict with earlier published 
foreign names for the same or different varieties, but which have been thor- 
oughly established through long usage, shall not be displaced unless long- 
used and available synonyms exist. 

3. Form of Names. The name of a variety shall consist of a single word. 

A. Varietal names shall be short, simple, distinctive, and easily spelled and 
pronounced. 

B. A varietal name derived from a personal or geographical name should 
be spelled and pronounced in accordance with the rules governing in the case 
of the original name. 

C. The name borne by an imported foreign variety should be retained, sub- 
ject only to such modification as is necessary to conform it to this code. 

I). Tin- name of a person should not be used as a varietal name during his 
lifetime. The name of a deceased person should not be so used except by 
the official action of this or other competent agronomic bodies. Personal 
names in the possessive form arc inadmissible. 

EL Nam< oi Itationf, States, or countries, in either the nounal or adjecti- 
val form should not be used as varietal nanus. 

I Such general terms as hybrid, selection, pure-line, pedigreed, seedling, 
etc., should not be used as varietal names. 

\ number, either alone or attached to a word, should not be used as a 

varietal Dame, but considered as a temporary designation while the variety 
is undcrKoinK preliminary testing. 

J I Nairn v. IikIi palpably cx;i^ci ate the merits of a variety shall be inad- 
missible. 

I In applying the provisions of this rule to varietal names which have be- 

mc firmly established in agronomic literature through long usage, no change 

shall l»e made which will involve loss of identity. 



BALL & CLARK: NAMING WHEAT VARIETIES. 



93 



4. Publication. A varietal name is established by publication. Publication 
consist's (1) in the distribution of a printed description of the variety named, 
giving its distinguishing characters, or (2) in the publication of a new name 
for a variety properly described elsewhere, such publication to be made in any 
book, bulletin, circular, report, trade catalog, or periodical, provided the same 
bears the date of issue and is distributed generally among agronomists and 
crop growers; or (3) in certain cases the general recognition of a name for a 
commercial variety in a community for a number of years may be held to con- 
stitute publication. 

A. Where two or more admissible names are given to the same variety, 
in the same publication, that which stands first shall have precedence. 

5. Citation. In the full and formal citation of a varietal name, the name 
of the author who first published it shall be given. 

6. Revision. No properly published varietal name shall be changed for any 
reason except conflict with this code, nor shall another variety be substituted 
for that originally described thereunder. 

Explanatory Comments on the Rules. 

The first clause (a) of Rule 1 will prevent the recognition of 
several different names for the same variety. Clause b permits the 
naming of pure-line selections, hybrids, etc., which have superior 
merit, even though not distinguishable by external characters. 

Rule 2 will govern the use of such names as Bluestem, Red Rus- 
sian, etc., when applied to two or more different varieties. Para- 
graph C provides against confusion which would result from com- 
pletely discarding well-known names. 

Rule 3 governs the formation of acceptable names. Canadian and 
Australian wheat breeders have set a splendid example in the applica- 
tion of short, simple, and appropriate names to the varieties they 
originate. Names like Huron, Marquis, Prelude, Preston, Pioneer, 
and Stanley in Canada, or Bobs, Comeback, Federation, Firbank, and 
Warren in Australia leave nothing to be desired. The various ex- 
planatory paragraphs, A to I, inclusive, show how the rule is to be 
applied in special cases. The ultimate effect will be to do away 
with long, cumbersome, and oftentimes misleading descriptive phrases 
and selection numbers now used as names. Paragraph I prevents 
confusion through the complete loss of familiar names of long 
standing. 

Rule 4 provides for the proper publication of varietal names. 
Williams 2 has given an admirable example of this in publishing his 

2 Williams, C. G. Wheat experiments. Ohio Agr. Expt. Sta. Bui. 298: 465- 
466. May, 1916. 



94 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



three - new varieties, Gladden, Portage, and Trumbull. Fuller de- 
scriptions will be desirable in official publication of new varieties. 

Rule 5 is of little consequence and can be eliminated without serious 
loss. It merely gives some credit to the author of a varietal name 
when discussing a variety in any formal or important connection, such 
as description or classification, or in an alphabetical checklist of all 
known varieties. Some of the varieties mentioned above would then 
appear as follows: Marquis (Saunders) ; Bobs (Farrer) ; Monad and 
Buford (Ball & Clark) ; 3 Portage (Williams), etc. 

Rule 6 governs changes of varietal names. 

Copies of this proposed code have been sent to Prof. E. G. Mont- 
gomery, chairman, Prof. C. G. Williams, and Dr. H. K. Hayes, com- 
prising the committee on varietal nomenclature of this Society. It is 
hoped that their discussion of it may be presented here before final 
action is taken. 4 



AGRONOMIC AFFAIRS. 

ANNUAL DUES FOR 1918. 

Those who have not already paid their dues for 1918 are urged 
to send checks promptly to the Secretary-Treasurer, P. V. Cardon, 
U. S. Department of Agriculture, Washington, D. C. Prompt remit- 
tance saves the Secretary-Treasurer much correspondence and insures 
continuous delivery of the Joi rxal. Under the by-laws of the 
Society, the JorKNAi. is not to be sent after April 1 to those whose 
due* are not paid before that time. The sending of back numbers 
entails extra work on the officers and adds materially to the So- 
iel 1 pense account. If your dues are not already paid, remit 
now and get each number as it appears. Don't forget that the 
amount i^ S2.50. And be sure to notify the Secretary of any change 
of address. 

*Ball, Carlcton R., and Clark, J. Allen. Bxperimentl with durum wheat. 
I ' - I '« p' Vr . I'.ul. fuK: 44. 46. n,iK. 

* The code, as here proposed, was adopted in its entirety, together with some 
i •■■;"■"! I--. tin < ..niiiiitiec. It is published in the report of the com- 
mitter < i< A m ik So< . A'.kON., o: 4-5 4^7. December, 1017). 



AGRONOMIC AFFAIRS. 



95 



THE SOCIETY'S HONOR ROLL. 



For some time, the Editor has planned to begin the publication of 
an honor roll of those members of the American Society of Agron- 
omy who are serving their country in the world war. Following is 
the list of those who are known to the Editor to be in the military 
forces of the United States or of Canada. No doubt there are 
many more. If you know of some member of the Society in the 
army or navy or engaged in war work whose name is not on the 
roll here printed, inform the Editor and the addition will be made. 



H. R. Cates, 
A. D. Ellison, 
Samuel D. Gray, 
P. H. Kime, 



Roll of Honor. 

E. E. Graham, 
Leroy Moo maw, 
J. V. Qutgley, 
Geo. T. Ratliffe, 
L. C. Raymond, 



F. J. Schxeiderhan, 
W. R. Schoonover, 
Herschel Scott, 
Paul Tabor. 



MEMBERSHIP CHANGES. 



The membership reported in the January issue was 653. Since 
that time 6 members have resigned and 9 new members have been 
added, a net increase of 3 and a total membership of 656. The 
names and addresses of the new members, names of the members 
resigned, and such changes of address as have come to the notice of 
the Secretary-Treasurer are reported below. 



New Members. 

Earl Burtis, 325 East Olive St., Fort Collins, Colo. 
Bruce J. Firkins, Dept. of Soils, I. S. C, Ames. Iowa. 
Alex. Granowsky, 320 Plum St., Fort Collins, Colo. 
Jerome Igo, 228 W. Magnolia St., Fort Collins, Colo. 
Fred Maier, 400 S. Howes St., Fort Collins, Colo. 
Sterling Minor, 318 W. Magnolia St., Fort Collins, Colo. 
Glenn Paxtox, Box 269, Fort Collins, Colo. 
Nelson S. Smith, School of Agr., Olds, Alberta, Canada. 
R. E. Stephenson, Dept. of Soils, I. S. C, Ames, Iowa. 



Members Resigned. 



Cobb, J. Stanley, Jenson, Chas. A., Wood, M. W., 

Garlaxp, J. J., Potter, R. S., Zerban, F. W. 



96 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Changes of Addresses. 

Abell, M. F., College of Agriculture, Storrs, Conn. 

Curtis. H. P., County Agent, Sutton, W. Va. 

Dunxewald, T. ]., 210 N. Carrol St., Madison, Wis. 

Goddard, L. H., 1324 Monroe St., Washington, D. C. 

Luckett. J. D., States Relations Service, U. S. Dept. Agr., Wash. 

Maris, Edwin I.. Demonstration Agent, Atwood, Kans. 

Warburtox, C. W., 320 Flour Exchange, Minneapolis, Minn. 

NOTES AND NEWS. 

George F. Corson, formerly professor of agriculture at the Iowa 
State Teachers' College, has been appointed assistant in soil survey 
at the Iowa station. 

H. B. Derr, for the past several years county agent in Scott Co., 
Mo., is now county agent in Fairfax Co., Va., with headquarters at 
Fairfax. 

R. A. Dutcher, formerly assistant professor of agricultural chem- 
istry at the Oregon Agricultural College, and C. A. Morrow, formerly 
professor of chemistry at Nebraska Wesleyan University, are now 
assistant professors of biochemistry in the Minnesota college. 

A. D. Ellison, for the past two years in charge of the cereal experi- 
ments on the U. S. Department of Agriculture's Arlington Farm near 
Washington, D. C, is now with the gas defense service of the U. S. 
Army. 

J. N. Else has been appointed assistant in agronomy in the Penn- 
sylvania college and station. 

\ . 1). Faville, animal husbandman of the Wyoming station, has been 
elected director of the station, succeeding H. G. Knight. 

R. L. Furry, a graduate of the Missouri College of Agriculture, is 
tant plant breeder on the Ferguson Seed Farms, at Sherman, 

Texas. 

S. ( . Harmon i assistant agronomist at the Virginia station. 

K. I . Holland, formerly county agent in Kimball Co., Ncbr., has 
laoY ;i 1 tant emergency county agent leader and has been suc- 
'1 in Kimball ( ounty by I'aul II. Stewart, instructor in agronomy 

in the Nebraska college last year. 

H. 1 ,. Knight, for the. past several years director of the Wyoming 
in and dean of the college of agriculture, lias been elected to a 
similar po ition in the ( >l<lahoma college and station. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. io. March, 1918. No. 3. 



THE EFFECT OF CERTAIN FACTORS ON THE CARBON- 
DIOXIDE CONTENT OF SOIL AIR. 1 

J. A. BlZZELL AND T. L. LYON. 

Review of Literature. 

It has been suggested 2 that some higher plants exert an influence 
on certain bacterial processes in the soil. Experiments indicate 
that some higher plants may, during the most active period of their 
growth, stimulate the formation of nitrates, while during the later 
periods of growth the same plants may exert a depressing effect. 
Since the conditions favoring the formation of carbon dioxide in soils 
are similar to those favoring nitrification, it is logical to suppose that 
the two processes would parallel each other. 

Russell and Appleyard 3 have recently produced evidence to show 
that, so far as the effects of temperature, moisture, and apparently 
aeration are concerned, such a correlation probably exists. The pres- 
ence of the growing plant, however, introduces so many far-reaching 
and disturbing factors as to make it difficult to institute such com- 
parisons when based on the ordinary methods of analysis. In the first 
place, carbon dioxide is produced not only by simple oxidation but 
also by plants, while nitrates are absorbed by plants. Again, under 

1 Contribution from the Laboratory of Soil Technology, College of Agri- 
culture, Cornell University, Ithaca, N. Y. Received for publication October 
23, 1917. 

2 Lyon, T. L., and Bizzell, J. A. Some relations of certain higher plants to 
the formation of nitrates in soils. Cornell Univ. Agr. Expt. Sta. Memoir No. 
1. 1913. 

3 Russell, E. J., and Appleyard, A. The influence of soil conditions on the 
decomposition of organic matter in the soil. In Jour. Agr. Sci., 8: 385-417. 
1917. 

07 



9 8 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



soil conditions the carbon dioxide produced is distributed between the 
liquid and gaseous phases. As the relative amount contained in each 
case is dependent on the content of water, calcium carbonate, and 
dissolved salts, it becomes difficult to determine quantitatively the 
effect of any one factor on the amount of carbon dioxide produced. 
It has frequently been observed that carbon dioxide is excreted by 
the roots of plants and that this process is in some way closely con- 
nected with the period of greatest vegetative growth. That there may 
be more deep-seated effects is indicated by results obtained by the 
authors. Boussingault and Lewy, Pettenkofer, Foder, Moller, Eber- 
mayer, and other early investigators observed that soil air contains 
more carbon dioxide than atmospheric air. This phenomenon was 
attributed generally to the oxidation of organic matter. The fluc- 
tuations in the carbon-dioxide content of the soil air were apparently 
governed by seasonal conditions. The effect of cropping appears to 
have been first studied by Wollny, 4 who placed calcareous sandy soil 
in metal cylinders and determined the carbon dioxide in the air once 
each week during the summer and part of the winter months. Com- 
paring grass sod with bare soil, he found less carbon dioxide under 
sod during the summer months and more during the winter months. 
He attributes the effect of the crop to its effect on the moisture, tem- 
perature, and porosity of the soil. 

Dcherain and de Moussy"' called attention to the fact that carbon- 
dioxide formation in the bare soil is due almost wholly to bacteria. 
They found that sterile soil at ordinary temperatures produces little 
carbon dioxide but that at points much above 65 C. considerable 
oxidation occurs by purely chemical means. They obtained large 
quantities of carbon dioxide by spreading soil in very thin layers and 
concluded that aeration is one of the most important factors. 

Molisch,* in studying root secretions, found an enzyme which lias 
the power to oxidize the organic compounds of humus. Czapek, 7 
working along similar lines, produced evidence to show that the plant* 
I'oui ' '-.en-lion which gives the acid reaction is carbonic acid. Wollny 8 

•Wollny. E. Untersuchttngen ubcr den Einfluss der Pflanzendecke und 

6a I'' chmttUflg anf detn KoMensanrcKdiall der Horicn Luft. In Forsch. 

Geb. AgriloPbysilc, 3: 1-15. iKKo. 

I Mm rain. IV I'., and tic Moussy, K. Snr I'Oxydation tie la Mature ()r- 
faniqtM 'In SoL /" Ann. AgRMB., 22 : 305-337. 1K06. 

' Moliscli II n,. r Wurzclausschied mid dercn Rinwirkung anf Or^anisclie 

tinien. tn Sitftmgi Akad Witt. Wien-Math. Nat., 96: 8.4-100. 1888. 

r CiapckfF. /nr Lchrt von der WttrxelaUliCheidnnKen. In Jalir. Wiss. Hot., 

29: 324. 

' Wollny, I \h< /< r ctzmtK der ( )rKanisclicn Stoffc. 1807. 



BIZZELL & LYON : CARBON DIOXIDE IN SOIL AIR. 



99 



demonstrated that, in the absence of free oxygen, organic matter may 
reduce the oxides of manganese and iron and form carbon dioxide. 
He also states that certain organic substances may form carbon 
dioxide by simple decomposition. Kossowitsch 1 ' grew mustard in 
nutrient solution mixed with washed quartz. He dissolved the carbon 
dioxide produced by percolating nutrient solution through the con- 
tainers. This was so regulated as to give 5 liters of percolate in 24 
hours. A check container on which no plants were growing was in- 
cluded. The carbon dioxide in the planted mixture increased grad- 
ually up to the end of the experiment when the plant was in full 
bloom, while the check varied within narrow limits. 

Stoklasa and Ernst 10 grew barley, wheat, rye, and oats in nutrient 
solutions and determined the amounts of carbon dioxide produced at 
different stages of growth. The younger the plants and more tender 
the roots, the greater the quantities of carbon dioxide produced per 
gram of dry matter. Considering the total amounts of carbon dioxide 
produced, however, they found the maximum with plants 70 to 80 
days old. At 84 days there were somewhat smaller quantities than 
at 80 days. They determined the quantity of carbon dioxide given 
off as gas by soil during 200 days. As the soil was bare of vegetation 
the action was attributed to bacteria. They obtained something more 
than twice the quantity of carbon dioxide estimated to be produced 
by wheat during 100 days. In a later article, 11 they investigated the 
chemical nature of root secretions and found carbon dioxide to be the 
principal one. 

Amberson 12 observed that the mucilaginous covering of the root 
hairs contains a saturated solution of carbon dioxide. Lau 13 deter- 
mined the carbon-dioxide content of soil air by a modification of the 
Petterson-Palmquist apparatus. He found that plant-root respiration 
has a marked effect upon the amount of carbon dioxide. It increased 
with the growth of the plant, reaching a maximum at the blooming 

9 Kossowitsch, P. The quantitative determination of carbon dioxide pro- 
duced by the roots of plants during the period of their development. In Jour. 
Expt. Agr. (Russia), 5: 482-493. 1904. Translation by J. Davidson. 

10 Stoklasa, J., and Ernst, A. Uber den Ursprung die Menge und die Be- 
deutung des Kohlendioxyds im Boden. In Centbl. Bakt., II, Abt. 14, S. 723- 
736. 1905. 

11 Stoklasa, J., and Ernst, A. Beitrage zur Losung der Frage der Chem- 
ischen Natur des Wurzelseketes. In Jahrb. Wiss. Bot, 46: 55 -102 - I0 <>8. 

12 Amberson, J. H. Ein Beitrag zur Kentniss der Natur der Wurzelaus- 
scheidungen. In Jahrb. Wiss. Bot, 47: 41-56. 1909- 

13 Lau, E. Beitrage zur Kentniss der Zusammensetzung der im Acker- 
boden befindlichen Luft. Inaug. Diss. Rostock. 1906. 



IOO 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



period. Potatoes and lupines gave larger amounts than other crops. 
This was attributed to the fact that potatoes and legumes have a 
higher rate of respiration. The carbon-dioxide content in cropped 
plats reached a maximum at the root zone, while in uncropped soil 
the content increased directly with depth of sampling. 

Barakov 14 grew lupines, clover, barley, rye, wheat, peas, vetch, pota- 
toes, and sugar beets in lysimeters. Several different types of soil 
were used. Samples of soil air were collected by attaching retorts 
which had been exhausted to the tubes at the bottom of the lysimeters. 
The carbon dioxide was then determined by absorption in standard 
barium hydroxide solution. The author found that the maximum 
carbon-dioxide content as a rule coincided with the period of bloom- 
ing. Contrary to the opinions of others he considers that plants pro- 
duce much greater quantities of carbon dioxide in soil than do bac- 
teria. He compared two lysimeters on which clover was grown. 
Both were harvested on June 14, and immediately thereafter the car- 
bon-dioxide content decreased rapidly. In one lysimeter the clover 
sod was plowed under and the carbon dioxide continued to decrease. 
In the other lysimeter the clover was allowed to grow and the carbon 
dioxide increased to a maximum at the second blooming. The great- 
est quantity of carbon dioxide produced by bacteria was 6.9 mgr. per 
liter, while the highest produced by plants was 27.3 mgr. per liter. 
Barakov quotes the work of Souprounenko as showing that lysimeters 
bare of vegetation produced less carbon dioxide than did those on 
which millet was grown. 

Comparing the effect of soil type, Barakov found that the ability of 
a particular plant to produce carbon dioxide is greater the more fertile 
the soil, and concludes that the effect is due directly to the more active 
vegetative growth on the fertile soil. He found that although the 
respiration curve varies with different plants, the maximum carbon- 
dioxide production occurs at the time of maximum life activity of a 
plant. 

Van Suchtclen 1 '' mixed 6 grams of calcium oxide with 6 kilograms 
of -oil and found the carbon dioxide produced to be less in the limed 
-oil. < Mi the oilier hand, magnesium sulfate, ammonium sulfate, and 
Ittperphospliatc stimulated carbon-dioxide production. The cxper- 

1 Uarakov, I'. The carbon dioxide content of soils during different stages 
"i plants. In Jour. I'.xpt. Agron. (Russia), 11: 321-342. 1910. 
Translation by J. JJavidson. 

an Snflitelen, ]• . II. II. Pbcr die Mcssung der Lcbcnstatigkeit der 
rn Hakterien im linden dun h die Kohlensaurcprodnktion. In Centb. 
Ilakt . II. Abt. 28, S. 45-«9. I9IO. 



BIZZELL & LYON I CARBON DIOXIDE IN SOIL AIR. 



IOI 



iments were meager and no statement is made as to the time inter- 
vening between application of salts and determination of carbon 
dioxide. 

Stoklasa, 16 in making a study of bacterial action in soil, concluded 
that the greatest production of carbon dioxide occurs in neutral or 
slightly alkaline soil abundantly supplied with air and readily as- 
similable plant nutrients. 

Lemmerman and his associates 17 quote the work of earlier inves- 
tigators as showing decreased production of carbon dioxide in soil to 
which calcium oxide and calcium carbonate were added. Lemmer- 
man found that applications of o.i percent, 0.5 percent and 1 percent 
of quicklime decreased production of carbon dioxide for the first two 
weeks, while the 0.5 percent and 1 percent applications continued to 
have the same effect for eight weeks. They also used calcium car- 
bonate in quantities chemically equivalent to the quantities of calcium 
oxide. They found quantities corresponding to 0.1 percent CaO to 
be stimulative while the larger quantities gave a decrease in carbon 
dioxide. They conclude that the decrease with the quicklime was due 
to direct absorption of the gas produced with formation of calcium 
carbonate. They call attention to what they consider the errors that 
arise in experiments of this kind (1) when carbon dioxide only is 
determined and methane ignored, (2) when calcium oxide absorbs 
the carbon dioxide formed, and (3) when calcium carbonate is ap- 
plied to acid soil, causing evolution of the carbon dioxide from the 
carbonate added. 

Leather 18 studied carbon-dioxide production in soil by determining 
the total carbon dioxide present in the form of a gas and in solution 
as Ca(HC0 3 ) 2 . This was done by taking a small soil core 4 cm. by 
8 cm., placing in a suitable container, and removing the gas by means 
of a vacuum. The author maintains that determinations of the gas- 
eous phase only do not represent carbon-dioxide production. From 
data obtained on the solubility of carbon dioxide as Ca(HCO s ) 2 at 
different pressures, the author estimates that when the total carbon 
dioxide is less than 10 percent and the soil is not particularly dry it 

16 Stoklasa, J. Methoden zur Bestimmung der Atmungsintensitat der Bak- 
terien im Boden. In Zeit. Landw. Versuch. Oesterr., 14: 1243-1279. 1911. 

17 Lemmerman, O., Aso, K., Fischer, H.. and Fresenius, L. Untersuchungen 
iiber die Zersetzung der Kohlenstoffverbindungen Verschiedene organischer 
Substanzen im Boden, Speziele unter dem Einfluss von Kalk. In Landw. 
Jahrb., 41 : 217-256. 191 1. 

18 Leather, J. W. Soil gases. Mem. Dept. Agr. India., Chem. Ser., 4: 85- 
132. I9I5- 



102 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



is nearly all in solution. The author found larger quantities of car- 
bon dioxide in the neighborhood of roots of crops than in fallow land. 

Russell and Appleyard 19 determined the carbon dioxide content of 
soil air in samples obtained by simple aspiration through a tube placed 
about 6 inches below the surface. They found considerable fluc- 
tuation in the quantities of carbon dioxide. The variations were 
attributed principally to seasonal changes. From November to May 
the temperature and carbon dioxide curves almost coincide, but early 
in May they diverge and do not come together again until November. 
Plotting the rainfall for the week preceding date of sampling, a rather 
close relationship was brought out. The cropped plots showed higher 

Per\ " 
cent 




jj»e TZn* Ju'<* J"'<f Jv'y Ju/y At)g Aug. A&}. S^pt S&pt.Stpt Sept. 

21 23 J /<? 9 26 2 9 ib 23 30 6 '3 20 Z7 



\ 7 U: 13. Diagram showing carbon dioxide in air from unlimed Dunkirk 

cropped and bare, with the mean atmospheric temperature for the 

week preceding each analysis. 

quantities of carbon dioxide than the uncropped, but the authors think 

thil if due to indirect effects. Passing from a neutral to a sour soil 
there u;i- an increa e in carbon dioxide, but different species of 
plantl bowed aboul the same production when grown on the same 

soil. 

I' -11 I J. .ind Appley.ird, A. The atmosphere of the soil; its com- 
position ;md <;ium- of variation. In Jour A^r. Sci., 7: 1-44. 1015. 



BIZZELL & LYON I CARBON DIOXIDE IN SOIL AIR. 



103 



These authors, 20 in summarizing the results of three years' work, 
conclude that the principal factors in carbon-dioxide production in 
the order of their importance are temperature, moisture, dissolved 
oxygen, and the growing crop. They obtained increased quantities 
of carbon dioxide on cropped soil, the two maxima occurring in May 
and August. The latter was the date of ripening. The authors argue 
that since little root activity occurs at the ripening period, the produc- 
tion of carbon dioxide can not be referred wholly to respiration. The 
authors did not find a depressing effect of the crop on carbon-dioxide 
production. Fred and Hart 21 compared additions of sulfates and 
phosphates and found in general the latter to be more effective. 



Per 
cent 

co 2 




Fig. 14. Diagram showing carbon dioxide in air from limed Dunkirk clay 



loam, cropped and bare. 

Potter and Snyder 22 treated soil with calcium carbonate, sodium 
nitrate, and ammonium sulfate singly and in combination. The car- 

20 Russell, E J., and Appleyard, A. The influence of soil conditions on die 
decomposition of organic matter in the soil. In Jour. Agr. Sci., 8: 385-417. 
1917. 

21 Fred, E. B., and Hart, E. B. The comparative effect of phosphates and 
sulfates on soil bacteria. Wis. Agr. Expt. Sta. Research Bui. 35- 1915- 

22 Potter, R. S., and Snyder, R. S. Carbon and nitrogen changes in soil 
variously treated. In Soil Sci., 1 : 76-94- I 9 I 6. 



104 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

bonate increased the production of carbon dioxide, but the other 
materials did not. The effect of the calcium carbonate disappeared 
after 59 days. 

The lack of uniformity in the results cited is no doubt due in large 
measure to the difference in conditions of the experiments and in the 
methods of determining the amounts of carbon dioxide produced. 
The evidence in the main points to the conclusion that fairly large 
quantities of carbon dioxide are excreted by plant roots, that this pro- 
duction reaches a maximum at the period of greatest vital activities, 
viz. the blooming period, and that it is increased by any factor which 
increases the vigor of growth of the plant. 

Regarding the effect of lime and the reaction of the soil on carbon- 
dioxide formation, there seems to be little unanimity of opinion. The 
discrepancy may be due, as Lemmerman points out, to the absorp- 
tion of the gaseous carbon dioxide when quicklime is applied and to 
the production of carbon dioxide by purely chemical means when cal- 

&er\ 

cent 
CO, 

<4 0- 
3.8- 




loatn, cropped and hare. 



cium carbonate i added to acid soils. • In cither case the determina- 
tion of carbon dioxide in the oil air by the methods described would 
not be a measure of carbOfl dioxide production. In the experiments 



BIZZELL & LYON ' CARBON DIOXIDE IN SOIL AIR. 



105 



to be described the results merely show fluctuations and not total 
carbon-dioxide production. These fluctuations were affected by so 
many uncontrolled factors that they do not necessarily parallel the 
production curves. However, interpreted in the light of these dis- 
turbances, the results show some interesting tendencies. 

Methods. 

The samples of air were collected from the drainage tubes at the 
bottoms of large lysimeter tanks, a description of which has already 
been published. 23 Each tank is slightly over 4 feet square and 4 feet 



Per 
cerri 




Juhe J^ne Jot/y Ju/y Ju/u Ju/u Aug. Aug. Aug. ^bq. Atgp. Sept.Segji Segt. 

Fig. 16. Diagram showing carbon dioxide in air from limed Volusia silt loam, 

cropped and bare. 

deep, with a capacity of about 3.5 tons of soil. They receive the 
natural rainfall but no other supply of moisture. Some of these tanks 
(1-12) were filled with Dunkirk clay loam soil in 1909 and the re- 

23 Lyon, T. L. Tanks for soil investigation at Cornell University. Science, 
n. s., 29 : 621-623. 1909. 



io6 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



mainder with Volusia silt loam in 1910 (tanks 13-16) and in 191 3 
(tanks 20-21). 

In collecting a sample of soil air a calibrated 500-c.c. Ehrlenmeyer 
flask was fitted with a 2-hole rubber stopper carrying- in each hole a 
Geissler stopcock. One of these stopcocks was connected by means 
of rubber tubing to a 15-liter aspirator bottle. The other was con- 
nected in a similar way to a brass Y-tube and the latter to the drain- 
age tube of the lysimeter. The remaining lower end of the Y-tube 
was connected to a rubber tube the lower end of which dipped under 
water. By this means it was possible to aspirate the air from the 



Per 
Cen1 

co 2 




Juno JunvJu/cj Ju/u Ju/u Jc//qAuq. Aug Aug. Auq. Acq- Sept Sept. Sent Sept. 

zi >>t) 5 7 a? 7 t9 a 7 sr aj 36 » /S 26 z7 

Pig. 17. Diagram ihowing carbon dioxide in air from uncropped Dunkirk 
clay loam, limed and unlimed. 

I3 imeter without interfering with the flow of drainage water. Soil 
air vrai drawn through the $<><> c.c. calibrated flask until the atmos- 

pherir air in the latter was entirely displaced. It was found that 
emptying tlx- aspirator bottle once was sufficient for this purpose. 

The As k wai then disconnected, taken to the head house adjoining, 

ISld Allowed to fand for a few minutes until the sample had risen to 
room temperature. The excesi pre lure inside the flask was relieved 



BIZZELL & LYON : CARBON DIOXIDE IN SOIL AIR. 



107 



by opening one of the stopcocks for a moment. Excess of standard 
barium-hydroxide solution was then added and the flask allowed to 
stand for 20 minutes with occasional shaking. The excess barium 
hydroxide was then determined by titration with standard oxalic acid 
solution. 

Effect of Crop on the Carbon-Dioxide Content of Soil Air. 

Determinations of carbon dioxide were made weekly from June 
21 to September 27, 1916, in samples of air drawn from tanks 3, 4, 7, 




Fig. 18. Diagram showing carbon dioxide in air from uncropped Volusia silt 
loam, limed and unlimed. 

8, 13, 14, 15, and 16, and the results plotted in curves shown in figures 
13 to 16, inclusive. Tanks 7 and 8 received applications of 3,000 
pounds per acre of quicklime in 1910 and 1915. Tanks 15 and 16 
received one application of 3,000 pounds per acre of quicklime in 



108 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

1913. Tanks 3 and 7 had been previously cropped as follows: 1910, 
corn; 191 1, oats; 1912, wheat; 1913 and 1914, timothy; 1915, corn. 
Curves 1 and 2 admit of a comparison between Dunkirk clay loam 
cropped to oats in 1916 (tanks 3 and 7) and the same type of soil 
which had been kept bare of vegetation since it was put in place in 
1909 (tanks 4 and 8). On June 21 the planted and unplanted tanks 
showed wide differences in the carbon-dioxide content. The cropped 
tanks showed a rapid increase up to the period of blooming, July 
12-19, an d then a rapid decline. The bare tanks showed a gradual 
increase, following in general seasonal variations and not reaching 
their maxima until August 30 to September 13. The fluctuations in 
the bare soil may be taken as representing bacterial action. After the 
maxima were reached there was a rapid decline both in the cropped 
and uncropped tanks. 

As the ordinary diffusion of carbon dioxide from the soil with a 
lowering of percentage obtains at all times, a fall in the crop curve 
would result during a period of nonproductivity and hence would not 
necessarily represent an effect on the bacterial production. In this 
decline should the cropped soil curve fall considerably below the bare 
soil curve, the results are to be interpreted as showing some inter- 
ference with bacterial activity. This is exactly what happened after 
August 2. On September 13, analyses of the air from the Dunkirk 
clay loam limed soil' showed that the air from the cropped tank con- 
tained 0.75 percent carbon dioxide while that from the bare tank gave 
4.2 percent. On August 30, the unlimcd cropped soil showed a car- 
bon-dioxide content of 0.85 percent, while the bare tank showed 4.3 
percent. It may be objected that as the determinations of carbon 
dioxide represent that in the gaseous phase only and as the cropped 
soil undoubtedly contained smaller quantities of water, the larger 
amounts of carbon dioxide found would be due to a relatively smaller 
quantity in the liquid phase and therefore would not be indicative of 
relative production. It was not considered advisable to disturb the 
soil in order to obtain samples for moisture determinations, but there 
were unquestionably smaller quantities of water in the cropped tanks, 
ai the following figure* of the total drainage from these tanks from 
June 5 to November 1, 1916, show. 

Tank 3, cropped 102.4 liters 

Tank 7, cropped 94.4 liters 

Tank 4, hare 306.0 liters 

Tank 8, bare 280.0 liters 



N : objection mi^ht apply therefore to some extent to the figures 



BIZZELL & LYON '. CARBON DIOXIDE IN SOIL AIR. 



109 



obtained prior to August 2, but on subsequent dates when the cropped 
tanks showed less carbon dioxide this factor could not have been pre- 
dominant. In fact, if the smaller percentages subsequent to August 2 
were influenced by the amount of moisture the depression due to 
cropping would be greater than is indicated by the curves. 

Figures 15 and 16 permit a similar comparison on Volusia silt loam 
soil. Tanks 13, 14, 15, and 16 were filled in 1910 and had therefore 
been in place approximately the same period of years as had the Dun- 
kirk clay loam. Tanks 14 and 16 had been kept bare of vegetation 
since being filled. Tanks 13 and 15 had been previously cropped as 
follows: 1913, oats; 1914, Canada field peas; 1915, corn. Comparing 
the cropped with the uncropped tanks, the curves are almost identical, 
markedly different from those .obtained with the Dunkirk clay loam. 
It appears that the crop had little effect on the carbon-dioxide con- 

Per ~~ 

cent 



CO, 




Fig. 19. Diagram showing carbon dioxide in air from cropped Dunkirk clay 
loam, limed and unlimed. 



tent of Volusia silt loam. It is significant that crop growth on the 
Volusia silt loam was considerably less than on Dunkirk clay loam 
and it may be that the difference in carbon-dioxide production is in 



I 10 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



some way connected with the vegetative activity of the plant. This 
idea is supported to some extent by the similarity between the curves 
of the two bare soils. 

Effect of Quicklime on the Carbon Dioxide in Soil Air. 

As previously stated, tank 8 received an application of 3,000 pounds 
per acre of quicklime in 1910 and a similar application in 1915. Tank 
16 received 3,000 pounds per acre of quicklime in 191 3. It is safe to 
assume that at the beginning of the season of 1916 all quicklime had 
been converted into calcium carbonate. The lime requirement (ac- 
cording to the Veitch method) of the surface foot of these soils before 
being placed in the tanks was approximately 3,000 pounds per acre. 
The soil in tanks 4 and 14 had never received an application of lime 
since being placed in the tanks. 

Tanks 4 and 8 received 10 tons per acre of manure in 1910 and 
tanks 14 and 16 had the same quantity in 1913. By reference to 
figure 17, it appears that liming Dunkirk clay loam had little effect on 



far 
cem 



CO, 




6 /,/ ,.J, , //,a> A'y/y yJ/y juiij /iAij. wJ</. "^ua^uq. 4u<y. 5vpt ScptSept. ^cpt. 

g f ga 5 /£ t9 a v <*> *3 •*<> 6 A3 *o Sit 

I i' .'n l>i;i«ram showing carbon dioxide 111 air from cropped Volusia silt 
I". mi, limed ;in<] milimcd. 



BIZZELL & LYON : CARBON DIOXIDE IN SOIL AIR. 



I I I 



the carbon-dioxide content. The figures seem to indicate the usual 
seasonal fluctuations. The effect of liming on Volusia silt loam is 
quite marked (fig. 18). The curves for the limed and unlimed soils 
are approximately parallel at all periods, but the liming apparently 
had a markedly stimulating effect on bacterial activity. The same 
general relations are seen when the figures for the cropped tanks 
limed and unlimed are plotted (figures 19 and 20). 

Liming did not increase the carbon dioxide in the air of cropped 
tanks of the Dunkirk clay loam, but had a decided effect on similarly 
treated tanks of the Volusia silt loam. It is significant that liming 
increased crop growth on the latter soil, while it had little effect on 
the Dunkirk clay loam. Volusia silt loam is a very heavy, compact 
soil and it is probable that the beneficial effect of line on the carbon- 
dioxide content is due mainly to an alteration of the physical con- 
dition rather than to its effect on the reaction of the soil. 

Effect of Burnt Lime versus Limestone on the Carbon Dioxide of Soil Air. 

Tanks 20 and 24 were filled with Volusia silt loam in 191 3. On 
May 1, 1916, tank 20 received an application of 6,000 pounds per acre 
of quicklime and tank 24 received 12,000 pounds per acre of lime- 
stone, ground to pass a 50-mesh sieve. Oats were raised on both 

Per I — 1 




Fig. 21. Diagram showing carbon dioxide in air from cropped Volusia silt 
loam, limed with burnt lime and with ground limestone. 



112 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

tanks. Figure 21 shows the results of carbon-dioxide determinations. 
Burnt lime gave larger quantities of carbon dioxide during the course 
of the observations (July 20 to September 27, 1916). The burnt lime 
curve is approximately parallel to the limestone curve. Both curves 
are similar to the cropping curves already discussed in that there is a 
decline after the period of greatest activity of the plant. Whether 
the burnt lime stimulated the bacteria to greater activity or whether 
the larger amount of carbon dioxide was due to a stimulation of the 
crop is not evident from the figures. The fact that the burnt lime 
tank yielded 931 grams of oats as compared with 842 grams from the 
limestone treatment lends support to the idea that the effect was due 
to greater crop growth. 

Summary. 

1. On Dunkirk clay loam cropping with oats produced striking 
fluctuations in the carbon-dioxide content of the soil air. The great- 
est apparent production was at the blooming period. Subsequent to 
the blooming period there was a marked decrease in the amount of 
carbon dioxide and this decrease was apparently due to the depress- 
ing effect of the crop on production by bacterial action. On Volusia 
silt loam the crop apparently had little effect on the carbon-dioxide 
content. 

2. On Volusia silt loam addition of quicklime increased the amount 
of carbon dioxide in the soil air. This effect was noticed both on 
the cropped and uncropped tanks. On Dunkirk clay loam quicklime 
apparently produced no effect. 

3. Treatment of Volusia silt loam with burnt lime was accom- 
panied by larger production of carbon dioxide than was the treatment 
with a chemically equivalent quantity of ground limestone. 



SNYDER: WHEAT BREEDING IDEALS. II3 



WHEAT-BREEDING IDEALS. 1 

Harry Snyder. 

Wheat is the ideal bread cereal. The physical character of its pro- 
teins is such as to impart bread-making qualities, while the nature 
and variety of the amino acids of these proteins give the maximum 
food value. Wheat is worthy of the high position assigned it by Sir 
William Crookes in his presidential address before the British Asso- 
ciation for the Advancement of Science in 1898, as "the most sustain- 
ing food grain of the great Caucasian race." Any improvement that 
can be effected in wheat is of the greatest benefit to mankind. 

The early history of wheat is shrouded in mystery. Presumably 
somewhere in Mesopotamia where modern man was nurtured., wheat 
had its origin. It has been stated that if wheat were not seeded and 
garnered by man it would soon become extinct, since it cannot exist 
as a volunteer crop. It would seem that man, in recognizing the great 
value of wheat as food, had taken in hand its propagation, and that 
nature had intrusted to him its care since it has lost the power to fend 
for itself. 

Some of the greatest of men have taken a deep interest in wheat 
improvement. General Washington, while active in the war for inde- 
pendence, had time to think of wheat. In one of his letters he says : 
" The wheat from some of my plantations by one pair of steelyards 
will weigh upwards of 60 pounds, and better wheat than I now have 
I do not expect to make." Referring to this same wheat some years 
later he wrote : " Xo wheat that has ever yet fallen under my observa- 
tion exceeds the wheat which some years ago I cultivated extensively, 
but which from inattention during my absence of almost nine years 
from home, has got so mixed or degenerated as scarcely to retain any 
of its original characteristics properly." 

During its thousands of years of domestication, wheat must have 
undergone some changes in both botanical structure and chemical 
composition, and it would seem that its development should be con- 
tinued and that it be still farther improved. 

1 Presented at the tenth annual meeting of the American Society of Agron- 
omy, Washington, D. C, November 13, 1917. The author of this paper is 
chemist for the Russell-Miller Milling Co., Minneapolis, Minn. 



114 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Seed selection and production upon rich suitable soils with fa- 
vorable climatic conditions have heretofore been the chief factors for 
improving the crop. But other important ways beside seed selection 
and proper nutrition are now open to us for the improvement of 
wheat, such as the selection and breeding of the parent wheat stocks. 
Parenthood in wheat is a factor for either improvement or retrogres- 
sion, as it is in any life process. I feel too incompetent to attempt to 
discuss any of the general laws of heredity as they apply to wheat, 
but I wish to note certain characteristics or properties of wheat, 
which it would seem should receive more attention in wheat breeding. 

It is always desirable to have in mind definite ideas as to what is 
desired to achieve in plant breeding, but it is particularly so in the case 
of wheat. The wheat breeder aims to secure new wheats or to im- 
prove old strains along one or more of the following lines : 

1. Large yield. 

2. Stiffness of straw. 

3. Resistance to rust and other diseases. 

4. Early maturity so as to escape heat, frost, and rust damage. 

5. Improved bread-making qualities and higher food value. 

The importance of maintaining and improving the quality of wheat 
deserves particular attention from the wheat breeder. It is not accom- 
plishing enough to secure a new wheat showing a larger yield per 
acre, if it is at a material sacrifice of the quality of the crop, when it 
is possible to secure a gain in both yield and quality. As a scholastic 
proposition, yield and quality might seem antagonistic characteristics, 
so that you could not expect to secure one without some sacrifice of 
the other. Happily the work of the late Dr. Saunders in developing 
the Marquis wheat shows that it is possible to secure both quality 
and yield as well as early maturity and hardiness. Since it has been 
proven possible to accomplish such results, I believe more attention 
should be m'ven to improving the quality of wheat. 

I 'l' .i- vary as to what is meant by wheat quality, which is a more 
Of lesfl indefinite term difficult to state with mathematical exactness. 
Quality cannot he expressed in percentage figures of special cou- 
nts. Ii i- i!m- bread maker and the housewife who are the final 
judges. If tlic wheat is of such character that the flour milled from 
it can he made into good bread, then the wheat is of good quality. 
'I he miller and the baker can ait in developing latent wheat quality, 
but they cannol create or impart it when it does not exist. It is much 
the fame as with btttter, if it is pleasing in taste and appearance it is 
of good quality. Neither wheat quality nor butter quality is as yet 



SNYDER: WHEAT BREEDING IDEALS. 



115 



capable of being determined absolutely by chemical analysis. Al- 
though wheat quality is difficult to measure, certain physical charac- 
teristics can be taken as a general index of quality and they will be 
found helpful to the wheat breeder. For example, a high protein con- 
tent is very desirable provided the protein is in such forms as to 
impart the best physical characteristics for bread-making, such as a 
certain degree of plasticity of the gluten. Hence, quantity and qual- 
ity of the gluten are a helpful guide to the wheat breeder. Many 
times, however, the wheat crop is hampered in showing its capabilities 
as to protein production because of lack of available nitrogen or other 
plant food. Seasonable variations also sometimes handicap the wheat 
breeder in arriving at conclusions. Wheat that is of fair glutenous 
character one year may be starchy another year, and some years the 
strongest glutenous wheats develop starchy tendencies. While these 
variations and handicaps are known to exist and must be recognized, 
certain general principles will be found helpful and can be applied by 
the wheat breeder. 

Suppose the wheat breeder has before him two samples of wheat 
of similar general character that have been grown and observed 
under like conditions, and their performance records are quite the 
same. He wishes to give preference to the one having maximum 
quality. In one sample, A, there are 80 percent of dark amber cor- 
neous or glutenous kernels and 20 percent that tend to be light col- 
ored and starchy; tests show the glutenous kernels to contain 16 per- 
cent protein and the starchy ones 12 percent. In sample B there is 
approximately the same ratio of glutenous and starchy kernels, but 
the analyses show 14.5 and 11 percent of protein for the two types of 
kernels. Naturally the preference would be given to the selected 
kernels of sample A, as the type of harder, stronger, and presumably 
better wheat. The selection, however, should be made not only on 
the basis of the amount of protein, but due regard should be given 
also to the physical quality of the gluten. If the gluten from the 
hard, selected B wheat is of the requisite plasticity and of superior 
quality to that of A, then preference should be given to B, as a crop 
from such seed will give the better bread-making value. The rule for 
selection should be: Get all the protein or gluten you can in the wheat, 
provided the gluten is of such character as to impart maximum bread- 
making quality. A high gluten with rather poor bread-making value 
is a poor combination. To breed wheats with the view of securing 
the maximum protein without regard to bread-making value is not 
working along right lines. 



Il6 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



In some wheat-breeding work it has been the aim to secure im- 
munity from rust and other diseases. The usual theory is to select 
as parent stock the most promising wheats as to powers of rust re- 
sistance. In some cases a selected durum has been taken as one of 
the parents. Particular attention must then be given to the other 
parent, because durum does not impart desirable bread-making qual- 
ities. Rust resistance is also obtained indirectly by early maturity. 
Hence, an early-maturing variety of high quality is more desirable 
than a later rust-resistant one of poorer quality. It takes time to pro- 
duce a good wheat by breeding alone, hence more progress will be 
made when the time element is considered by seed selection from well- 
established varieties rather than by premature attempts to replace old, 
well-known types with new and questionable kinds. When a com- 
munity has established a reputation for producing a wheat of quality 
the maximum price with a special premium is paid for such wheat. 
When the wheat is of poor or mediocre quality its marketing is at- 
tended with difficulties unless there is a general wheat shortage, when 
a poor quality commands an abnormal price. This principle of not 
encouraging the introduction of a poorer wheat where a community 
has established a reputation for raising the best quality of wheat is 
recognized by Dr. Saunders in his recommendations in the distribution 
of his Preston and Marquis wheats. 2 

Preston, Huron, and Stanley, by careful selection, have been considerably 
improved ; however, they do not produce flour of the highest baking strength, 
a disadvantage the seriousness of which can easily be exaggerated, but should 
not be overlooked in those districts where wheat is grown for export, and 
where a reputation for remarkably high baking strength has already been 
established. This applies particularly to the central parts of Canada. . . . 
Taking all points into consideration, Marquis wheat is recommended as the 
most promising sort at present available for farmers who require a hard, red 
wheat of lii^li baking strength and ripening earlier than Red Fife. 

The production of an inferior quality article in a locality where a 
reputation has been established for superior goods is more serious 
than might appear at first consideration. About ICS70 the Mohawk 
Valley of New York produced a superior quality of cheese that com- 
manded a premium in the English markets. A few years later the 
doctrine "i combined cheese and butter production was promulgated. 

iL-ned thai the milk could he skimmed a little and the cheese 
still be ju ) a- good, while both cheese and butter could be made and 
lOld. Just about that time the Canadian cheese industry started, and 
' did not ikim, but made Cheese only. As a result the Mohawk 
a Saunders, C. EL In ( inadian Kxpt. barms Kpt., 1010, p. 171. 191 1. 



SNYDER : WHEAT BREEDING IDEALS. 



117 



Valley cheese lost its English market, which was won by Canadian 
full-milk cheese. Following this loss of the cheese market many 
farms declined in value from $10 to $20 per acre. It is sometimes 
difficult for an agricultural community to change its procedure and 
the loss of a good market for a staple product is a serious matter. A 
reputation for a quality product is a valuable asset for any community. 
After the war is over, and Kaiser " kultur " is discredited as it de- 
serves, wheat culture will return to normal conditions ; wheat will 
cease to be a war-stimulated industry and it will then become a world- 
competitive industry where quality will be an important commercial 
factor. 

In the promulgation of new wheats that are lacking in high bread- 
making qualities but otherwise possess desirable characteristics it is a 
mistake to argue that the miller is prejudiced against the wheat and 
that he is not willing to accept it at its face value. The miller, as a 
rule, is only too glad to get wheat of quality, and he is not liable to be 
mistaken in his judgment of a new wheat unless he mills and markets 
only a small quantity of it. A heavy-weight, good-appearing, but 
poor bread-making wheat may form a small part of a wheat milling 
blend and not show its poor quality as readily as would be the case 
if the new wheat alone were milled in quantity. It is a real test of a 
wheat-breeder's character when he recognizes the defects of a good- 
looking wheat that proves to be only a mediocre bread-maker. Hap- 
pily there are such scientists, as the quotation just cited from Dr. 
Saunders shows. 

In testing new varieties of wheat, the small 2-roll sample mill will 
be found helpful if properly used and attempts are not made to solve 
milling problems that are not capable of being approached with such 
meager facilities. A small, so-called experimental mill will enable a 
sample to be prepared for testing the amount and physical quality of 
the gluten, and in a general way to tell if the wheat is likely to yield 
a flour that is reasonably responsive to yeast action. If a wheat passes 
such a test then it has reached a point where it can be tried in a large 
way and enough produced for trial in a small but reasonably well- 
equipped commercial mill. The wheat will finally be measured by 
regular commercial standards and the sooner a wheat breeder has sub- 
mitted his new wheat to such a test the better position he is in to judge 
of the quality of his new product. To attempt to put out a new wheat 
with insufficient testing as to its quality or to cover up a defect is, to 
say the least, unscientific, as any shortcomings a wheat may possess 
are sure to be discovered when put to the test. 



Il8 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



The small sample mill cannot be used for determining the flour 
yield of a wheat or for some of the finer points in connection with 
flour qualities. There are no short cuts in flour milling, that is, no 
important operations can be abbreviated or omitted and at the same 
time a quality flour made. The final test of a wheat must be its de- 
portment under commercial conditions. I do not wish to discourage 
the use of the sample mill, but to encourage its proper use in wheat 
breeding. 

The bulletins and circulars issued by one of the experiment stations 
the past year give some interesting data on the comparative workings 
of the sample or experimental mill and a small commercial mill. 
Some of the same wheats were milled both experimentally and com- 
mercially. Calculations from the published data show that a bushel 
of the same wheat yielded on the sample mill 5 pounds more flour and 
25 percent less feed, with 2 pounds per hundred more of invisible loss 
than from the commercial mill where the flour was much superior. 
The sample mill fails to make a proper separation between the feed 
and the flour. When you check the workings of a small sample or 
experimental mill against a well-equipped commercial plant many 
amusing inconsistencies appear that would not be apparent to the 
layman. Flour milling is a high-grade mechanical industry, and the 
plant breeder should call upon the miller for expert assistance in this- 
line rather than attempt to become a miller in addition to being a plant 
breeder. Cooperation upon such questions is the best way to get 
results. 

There are many important side lines or problems which the wheat 
breeder can take up in connection with his regular work, and which 
have an important bearing upon the main line he is investigating. The 
tendency of wheats to become starchy or to develop what is sometimes 
called yellowberry is one of these, problems. While it appears to be 
due in part to climatic conditions, it is also due to lack of available 
plan! food, particularly nitrogen, as well as to other causes. 

For a t in' mum believed that wheats grown on irrigated lands 
were necessarily starchy in character, but of late years it has been 
found that wheal of very high quality can be raised by irrigation with 
proper control of the water supply. The problems of plant food, 
water BUDply, and quality of crop are open for further investigation. 

I'lKc'f.n fli r,-i r ;»ive the wheal breeder additional problems to 
contend with. The value of formaldehyde as a general fungicide 
appears to be well known, but many fanners have not as yet been suf- 
ficiently impressed with the fact thai it is possible to use so strong a 



SNYDER : WHEAT BREEDING IDEALS. 



119 



solution as to impair seed germination. During the past year my 
attention has been called to cases where large losses from overtreat- 
ment have resulted. It is not safe to conclude that if a definite 
strength of formaldehyde solution is good, a larger amount would be 
better. 

National grading will doubtless prove valuable in gradually raising 
the quality of wheat. When each State and market had different 
rules for grading, it not infrequently happened that a wheat receiv- 
ing a No. 2 grade in one market or State would get a No. 1 grade in 
another, or vice versa. This lack of uniformity has led to serious 
misunderstanding and trouble. The marketing of clean grain needs 
to be encouraged. More fanning mills should be used on farms, then 
less valuable transportation space would be required for marketing 
the cleaner wheat and the screenings could be fed to live stock on the 
farms. It never pays to market a dirty product and this applies par- 
ticularly to wheat. The national grading of grain must necessarily 
recognize the commercial characteristics of the grains graded, which 
cannot be based on an imperative academic basis. Any attempt to 
make it appear that wheat is always wheat and that all weights, kinds, 
and types are equally valuable for flour and bread-making purposes, 
simply encourages poor, slovenly methods of farming. 

The wheat breeder should encourage wheat farmers to produce and 
disseminate the best varieties of wheat. Seed-wheat farms where 
high-grade pedigreed wheat is produced are as necessary as stock 
farms where pedigreed stock is raised. By improving jointly the 
yield and quality of wheat the wheat breeder renders an important 
service to mankind. The best of bread-making wheats selected so as 
to have the maximum protein need little or no supplementing with 
other more expensive protein-containing foods. The best results are 
secured from a liberal mixed diet, though good bread that is rich in 
protein supplies at a minimum cost the maximum of nutrients. 



120 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



NATURAL CROSS-POLLINATION IN WHEAT. 1 

H. K. Hayes. 
Introduction. 

The pure-line method of breeding has been almost universally 
adopted for self-fertilized crops. The initial selection is generally 
considered simply to isolate those types which are present in the com- 
mercial variety. Continuous selection in pure lines of self-fertilized 
crops has not, as a rule, given any improvement of importance, al- 
though a few cases are on record of a sudden change in a pure line 
which proved to be of selection value. Almost all commercial varieties 
of farm crops are mixtures of various types. Even though all plants 
are of the same general appearance, selection often isolates types 
which differ in such important things as yield and resistance to 
lodging. 

As an illustration of the results of such selection, the production 
of Havnes Bluestem wheat, Minnesota No. 169, may be mentioned. 
Before the recent introduction of Marquis wheat, this pedigreed type 
was widely grown in Minnesota. Minnesota No. 169 was one of a 
large number of selections which were first tested at the Minnesota 
station in 1892. The initial selection was a single plant which gave 
progeny of promise and which, when tested in various parts of Minne- 
sota, gave a marked increase as compared with the commercial types 
then grown. 

Minnesota No. 169 is a hairy-chaffed, awnless wheat with red 
coloring matter in the bran layer of the kernel. An examination in 
[915 <>f bulk seed of this strain showed the presence of some unpig- 
mcnted or white kernels. These white kernels bred true to the Blue- 
stem type of plant for other characters. Several samples of Minne- 
\*o. [69 .1- grown by fanners were obtained in 1916 and were 

examined. Nearly all of these samples contained some white kernels 
which, when grown, proved to be similar to Naynes Bluestem in other 
characters. 

If Bltiettetn wheal was originally pute for the red color of the 
kernel, which seems very likely, the question at once arises as to 

iPtlMisJirfl with tlir approval of tlic I Erector as Paper No. 101 of the 
( •• ' ■ 1 * of tin M 111 ota Agricultural Kxpcrimcut Station. Received 

for publication hrcrmher 16, 1917. 



HAYES: CROSS-POLLINATION IN WHEAT. 



121 



whether the presence of white kernels is due to crossing or to a loss 
mutation. In either case the breeder is faced with the necessity of 
purifying the type and one naturally wonders how frequently it will 
be necessary to repurify a " pure " line. 

It seems reasonable to suppose that natural cross-pollination may 
be a frequent cause of the production of inherited variations within 
pure lines. If there is much cross-pollination, a careful examination 
of nursery plots would seem the best possible means for determining 
the frequency of the occurrence, for under these conditions many 
different types are being grown at short distances from each other. 

Previous Studies of Natural Cross-Pollination in Wheat. 

For a detailed discussion of studies in this field, the reader is re- 
ferred to a review made by Pope (1915). 2 Some of the more impor- 
tant work along this line may be briefly repeated. Nilsson-Ehle 
(191 5) was able to show by an experimental test that some varieties 
are much more liable to natural cross-pollination than others. 

Only a few cases of cross-pollination have been recorded in North 
America. In 1903 Saunders (1905) announced a single case of a 
natural hybrid. Smith (1912) found eight natural hybrids in 96 rows 
of Turkey winter wheat. More recently, Leighty (191 5) has de- 
scribed four cases of natural crosses between wheat and rye, and con- 
cludes that the existence of such hybrids indicates that wheat is more 
often cross-pollinated than was formerly supposed. 

Results at the Minnesota Station. 

In 191 5, nursery plots of Haynes Bluestem wheat were grown as 
checks for the purpose of determining the variability of the field used 
for the nursery work. Plots of 100 plants, placed in the form of a 
square, were used at this time. The Bluestem plots were placed every 
fifth plot and at maturity a number of individual plants were selected 
for the purpose of producing pure lines of Bluestem. 

Fifty of these plants were grown in 1916 in individual plots. Of 
these 50 types, 3 proved definitely that they were natural hybrids and 
gave ratios which were according to Mendelian expectation ; 1 gave 
45 brown and 19 white chaffed plants ; another gave 32 hairy brown 
chaff, 16 hairy white chaff, 13 smooth brown chaff, and 5 smooth 
white chaff plants ; the third gave 46 hairy and 19 smooth chaff plants. 
Two other of the Bluestem selections gave both red and white 
kernels. 

2 References are to " Literature cited," page 122. 



i 



122 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Forty-seven plant selections of Marquis were grown in 1916 from 
selections made in field variety test plots. Of these all were true for 
the Marquis type of head, but two gave both red and white kernels. 
Thirty- four other pure lines were grown from individual plants, no 
natural hybrids being observed. 

In the winter wheat nursery, 54 selections were carefully examined ; 
of these, one proved to be a natural hybrid and gave definite segre- 
gation ratios. 

Approximately ico individual plant selections of spring wheat were 
made in 1916 in the nursery and were grown in 1917. No segrega- 
tion occurred, although four plants were discovered which were be- 
lieved to be F 1 crosses. It is a comparatively easy matter to tell a 
first-generation cross between Marquis or Bluestem and any bearded 
variety of wheat, because the first generation has awns which are in- 
termediate in length. 

Thirty-six selections of types which were of promise in 1916 were 
tested in rod rows in 191 7. Two rod rows of each selection were 
carefully examined to determine any impurities. Of these 36 plots, 
21 had from one to eight plants that were clearly first-generation 
crosses. 

Conclusion. 

The results here reported would indicate that conditions in 191 5 
and 1916 at University Farm were either very favorable for natural 
crossing in wheat or that its occurrence is much more frequent than 
has been generally supposed to be the case. 

Literature Cited. 

Leighty, C. E. 

Natural wheat-rye hybrids. In Jour. Amer. Soc. Agron., 7 : 209-216. 

Nilsson-Khee, H. 

1915. Gibt es erbliche Weizenrassen mit inelir odor wcniger vollstandigcr 

Selbstbefruchtung? In Ztschr. Pflanzenzucht, Bd. in, Heft 1, 
S. 1-6 

Pope, M. N. 

1916. I Ik mode of pollination in some farm crops. In Jour. Amer. Soc. 

Agron., 8: 209-227. 
Saunders, C. E. 

I'/j". A natural hybrid in wheat. In Proc. Amer. Breeders' Assoc., I : 
I37-I3& 
Smith. L H. 

I'/i.v Oirtirrnur of natural hybrids in wheat. /;/ Proc. Amer. Breeders' 

Assoc., v. 7-8, p. 412-414. 



HAYES : SELF-FERTILIZATION IN CORN. 



123 



NORMAL SELF-FERTILIZATION IN CORN. 1 

H. K. Hayes. 
Introduction. 

Methods of breeding self-fertilized plants have been standardized 
so that at the present time the same general plan is used by nearly all 
scientific investigators. Many minor points are not yet entirely 
settled, but these are of relatively little importance. With corn, how- 
ever, there is little uniformity of opinion among plant breeders as to 
the actual value of different methods of work. With regard to the 
correlation of various ear and plant characters and resultant yield of 
the progeny a large amount of data shows that there is no significant 
relation. This lack of correlation between ear and plant characters 
and yield of progeny in corn is partially explained by the fact that en- 
vironmental effects modify various ear and plant characters to a 
marked degree. The mode of pollination of the corn plant may be 
given as a second cause of this lack of relation. 

Studies by East (1908, 1909), 2 Shull (1908, 1910), Collins (1910), 
and others, together with further data, have been reviewed by East 
and Hayes (1912). One of the conclusions reached as a result of 
these investigations was that self-fertilization in plants that naturally 
cross-pollinate reduces vigor and cross-pollination in self-fertilized 
plants tends to increase vigor. It was believed that this phenomenon 
of increased vigor in first-generation crosses, which has been called 
heterosis by Shull, was a physiological stimulus due to heterozygosis, 
although it was recognized that some factors in the heterozygous con- 
dition gave a greater stimulus to development than others. A recent 
hypothesis of Jones (191 7) is very attractive and places the matter 
on a strictly inheritance basis. He attributes the increased vigor 
which is often obtained in the heterozygous condition to growth 
factors. This seems logical, as nearly all experiments show that the 
heterozygous condition for each particular growth factor gives more 
than half as great a result as the homozygous condition. While 
dominance is not complete, there is almost always a tendency to 

1 Published with the approval of the Director as Paper Xo. 100 of the Jour- 
nal Series of the Minnesota Agricultural Experiment Station. Received for 
publication December 16, 1917. 

2 References are to " Literature cited." p. 126. 



124 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



dominance. The hypothesis of linkage is used to explain why it is 
impossible to obtain all growth factors in a single homozygous in- 
dividual. 

Efffxts of Self-Fertilization on Yield in Corn. 

Previous investigators have shown that the first year of self- 
fertilization in corn often causes large reductions in vigor. As a con- 
tribution to the question of the actual amount of reduction in yield 
due to such fertilization the following comparison of the yields of 
normally pollinated Minnesota No. 13 yellow dent corn and 15 first 
generation self-fertilized lines is given. 



Table i. — Yield in bushels per acre of normally pollinated Minnesota No. 13 
corn and 15 first generation self-fertilised lines. 



Variety. 


Mode of 


Vield, 


Variety. 


Mode of 


Yield, 


pollination. 


bushels per acre. 


pollination. 


bushels per acre 


Minn. 13 


Normal 


49-3 


13-46 


I year selfed 


12.3 


13-9 


1 year selfed 


15-4 


13-31 


do. . 


25.2 


13-12 


do. 


14.8 


13-47 


do. 


25.8 


13-17 


do. 


34-2 


13-56 


do. 


29.4 


13-19 


do. 


18.7 


13-78 


do. 


28.4 


13-25 


do. 


20.1 


13-99 


do. 


7-9 


13-30 


do. 


43-1 


13-199 


do. 


24-3 


13-34 


do. 


I9.4 














13-35 


do. 


41.0 


Avprai'p splfprl lines 


24.0 







The normally pollinated Minnesota No. 13 gave a yield of 49.3 
bushels, while the average of the selfed lines gave a yield of 24.0 
bushels. This is slightly more than a 50-percent decrease in yield for 
the self- fertilized as compared with the normally pollinated seed. 

The amount of normal self-fertilization is, therefore, of interest. 
A preliminary study was outlined at Minnesota and started in 1916 
with the purpose of dete rmining the percentage of normal self-pollina- 
tion under Minnesota conditions. 

PttVXOUf Studies. 

Waller f 1 'y J 7 ) lias recently reported preliminary results as ob- 
tained at the Ohio station, lie used yellow and white dent varieties 

of com and planted bill- of three stalks each of Wing Uundrcd-Day 

White in .1 field of l\< id Yellow Dent. The bills of Wing white dent 

were placed at tome distance apart and before pollination two of the 

tbree tall from each bill of tbe white corn were detasseled. The 
pel ei '.!"< of white kerneb was then determined for each ear of each 
stalk of tbe wbite corn UOOfl whi< b a tassel was allowed to remain. 
There was found to be considerable difficulty in separating tbe wbite 



HAYES! SELF-FERTILIZATION IN CORN. I25 

and yellow seeds. An average of 5.13 percent of self-pollination was 
obtained from these tests. 

Results of this nature would be of almost enough value to pay all 
seedsmen and farmers to use seed from detasseled rows, thus in- 
suring cross-pollination. This method has been encouraged by the 
Illinois station (Hopkins et al., 1905). 

As Waller has pointed out, a number of experiments will need to 
be made before definite conclusions as to normal self-pollination can 
be reached. 

Results at the Minnesota Station. 

In the Minnesota test, Rustler white dent and Minnesota No. 13 
yellow dent were used. Single seeds of Rustler white dent were 
planted in hills at some distance from each other and a stake was 
placed beside each white seed. Due to various causes only six ears 
grown from the white seeds were obtained. The resultant progeny 
were then carefully examined and a separation was made into three 
groups, yellows, doubtful yellows, and whites. It was found to be 
very difficult to separate the seeds, although great care was made to 
place only those seeds with some yellow color in the endosperm in 
the yellow group. The result of this classification is given in Table 2. 



Table 2. — Yellow, doubtful yellow, and white seeds obtained from cars of 
Rustler white dent grown in a field of Minnesota No. 13 yellow dent corn. 



Ear No. 


Yellow. 


Doubtful yellow. 


White. 


'■■ 


395 


22 


6 


2 


335 


20 


7 


3 


179 


23 


16 


4 


396 


35 


10 


5 


36i 


. 101 


27 


6 


245 


28 


3 


Total 


1,911 


229 


69 



The following year a number of hills of the doubtful yellow and 
white groups were grown and approximately 25 ears from each group 
were artificially self-fertilized. All of these ears at maturity con- 
tained a considerable percentage of yellow seeds, which proves that 
nearly all seeds were cross-pollinated the previous year. 

If the classification had not been tested the whites would have been 
considered pure white and the doubtful yellow and yellows would 
have been considered yellow. This shows the impossibility of cor- 
rectly classifying seeds of crosses between yellow endosperm dent 
and white endosperm dent varieties of maize without determining the 
degree of error in the classification. 



126 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Conclusions. 

The results here reported appear to justify the following con- 
clusions : 

1. The first year of self-fertilization in maize, on the average, 
causes a reduction of about 50 percent in vigor as determined by the 
yield in bushels of shelled corn per acre of normally pollinated and 
self-fertilized strains of Minnesota No. 13 dent corn. 

2. If there is normally as high as 5 percent of self-fertilization 
under field conditions, it. might pay seedsmen and corn growers to use 
seed from detasseled stalks, thus using only cross-pollinated seed. 

3. The amount of normal self-pollination under field conditions as 
determined by a somewhat limited test proved to be certainly less 
than 5 percent. These results, although too few to base conclusions 
upon, are here given with the hope that other investigators will make 
similar tests. 

Literature Cited. 

Collins, G. N. 

1910. The value of first generation hybrids in corn. U. S. Dept. Agr., 
Bur. Plant Indus. Bui. 191, 45 p. 

East, E. M. 

1908. Inbreeding in corn. In Conn. Agr. Expt. Sta. Rpt. for 1907, p. 

419-428. 

1909. The distinction between development and heredity in inbreeding. 

In Amer. Nat, 43: 173-181. 
East, E. M., and Hayes, H. K. 

1912. Heterozygosis in evolution and in plant breeding. U. S. Dept. Agr., 
Bur. Plant Indus. Bui. 243, 58 p. 
Hopkins, C. G., Smith, L. H., and East, E. M. 

1905. Directions for breeding of corn. 111. Agr. Expt. Sta. Bui. 100. 
Jones, D. F. 

1917. Dominance of linked factors as a means of accounting for heterosis. 

In Genetics, 2: 466-479. 
Waller, A. E. 

i<>ij. A method for determining the percentage of self-pollination in 
maize. In Jonr. Amer. Soc. Agron., 9: 35-37. 



M'CALL & RICHARDS: FOOD REQUIREMENTS OF WHEAT. 127 



MINERAL FOOD REQUIREMENTS OF THE WHEAT PLANT AT 
DIFFERENT STAGES OF ITS DEVELOPMENT. 1 

A. G. McCall and P. E. Richards. 
Introduction. 

Two years ago in a brief paper published in the Journal of this 
Society 2 the senior author called attention to the desirability of a care- 
ful study, under controlled conditions, of the mineral food require- 
ments of the principal farm crops plants at different stages of their 
development, with a view to working out the fundamentals of a 
rational fertilizer practice. In this original publication and in a sub- 
sequent paper in Soil Science 3 the method of attack of this problem 
is described. In addition to describing the method, the last-men- 
tioned paper gives the results for the first 24-day growth period for 
wheat. 

In a paper read before the Society for the Promotion of Agricul- 
tural Science and printed in the Proceedings of that Society for 1916, 4 
the senior author has shown that the ratio of mineral nutrients giv- 
ing the best growth rate for this early period is practically the same 
for the soybean as for the wheat plant. During the past year the 
work has been extended to cover three stages in the development of 
the wheat plant ; namely, the first 30 days, the second 30 days, and 
finally the period extending from the close of the second 30-day 
period to the maturity of the plant. 

Method of Experimentation. 

The salts employed in making up the nutrient solutions were mono- 
potassium phosphate, calcium nitrate, and magnesium sulfate. 
Thirty-six different proportions of these salts were used, thus neces- 

1 Contribution from the Department of Soils of the Maryland Agricultural 
Experiment Station, College Park, Md. Presented by the senior author at 
the tenth annual meeting of the American Society of Agronomy, Washington, 
D. C, November 12, 1917. 

2 McCall, A. G. A new method for the study of plant nutrients in sand cul- 
tures. In Jour. Amer. Soc. Agron., 7: 249-252. 1915. 

3 McCall, A. G. The physiological balance of nutrient solutions for plants in 
sand cultures. In Soil Science, 2: 207-253. 1916. 

4 McCall, A. G. The physiological requirements of wheat and soybeans 
growing in sand media. In Proc. Soc. Prom. Agr. Sci. for 1916, p. 46-59. 1917. 



128 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



sitating the growing of 36 different cultures for each of the three 
growth periods. Since the dry weight was used as a measure of the 
growth rate during the different periods, it was of course impossible 
to carry through to maturity the same individuals. According to the 
plan adopted, the first series of cultures were grown for a period of 
30 days, when they were harvested and the dry weights determined. 
A second series of cultures were then started from the same lot of 
seed as that used for the first series. During the first 30 days these 
plants of the second series were given a nutrient solution containing 
the same proportion of the three salts as had been found to give the 
best growth rate during the first growth period. From the close of 
the first 30-day period to the end of 60 days the individual cultures 
of the second series were supplied with the same different proportions 
of the three salts as was employed in the first series. At the end of 
60 days the second series was harvested and from the dry weights the 
best proportion of salts for the second period was obtained. In a 
similar manner the plants for the third series were grown from the 
same lot of seed and supplied with the best salts proportions during 
the first and second periods, after which they were given the differ- 
ential feeding and harvested at maturity. 

The growth period for Series I extended from October 10 to No- 
vember II, 1916; Series II from December 18, 1916, to February 17, 
1917, and Series III from March 14 to June 25, 1917. Plates 2 and 3 
show the appearance of the plants at the end of these growth periods. 
The seed used was northwestern grown spring wheat of the Mar- 
quis variety obtained from the Olds Seed Co., Madison, Wis. Spring 
wheat has proven to be much more satisfactory than winter wheat, 
since under greenhouse conditions the former develops a stiffer straw 
and is less difficult to bring to maturity than the latter. 

The plants were grown in pure washed quartz sand, the culture pots 
being arranged on rotating tables, in order to obtain uniformity in ex- 
ternal environmental conditions. In the earlier work granite-ware 
pot! were used, bill during the past year clay pots have been em- 
ployed, in order to avoid bringing the nutrient solution into contact 
with ioldered joints. The form of the pot which is now being used 
and the method of supporting tbe plants are shown in Plate 3. Figure 
22 i a CTOI lection view of the pot and shows the arrangement of 
tli' -upply funnel and the outlet tubes. 

[fl order tO make the containers impervious t0 moisture and at the 
same time prevent the solution from coming into contact with the 
day, die pott irere first boiled in paraffin to drive the air out of the 
walls and then brushed Over both inside and outside with a thin coat 



Journal of the American Society of Agronomy. 



Plate 2. 




Fig. 2. Wheat plants at the close of the second 30-day period. The plants 
are supported by an open wire frame which permits the leaves to assume a 
natural position. 



of the American Society of Agronomy. 



Plate 3. 




U'h 



maturity. The screen at the 



•ear has one-inch 



M'CALL & RICHARDS: FOOD REQUIREMENTS OF WHEAT. 129 



of hot paraffin. The glass funnel (D) rests on an inverted glazed 
porcelain dish (F) which not only supports the funnel but also serves 
to insure a more thorough distribution of the solution throughout the 
mass of sand. The glass out- 
let tube is held in place by a 
paraffin seal (B) and has its 
inner end resting on the bottom 
of the pot near the center. The 
bell-shaped enlargement (H) at 
the end of this tube is closed 
against the entrance of sand by 
the insertion of a plug of glass 
wool. A short length of glass 
tubing inserted ahead of the 
wool effectively provides against 
the possibility of the plug being 
drawn out when suction is ap- 
plied for the removal of the 
solution. The substratum used 
in these cultures consisted of 
medium fine white sand which 
has been washed several times 
with distilled water from a 
Barnstead still. 




Fig. 22. Cross-section view of the clay 
pot used for sand cultures. A, paraffined 
clay wall ; B, paraffin seal ; C, wax seal 
covering surface of sand; D, supply fun- 
nel made by cutting bottom out of a 
wide-mouthed bottle ; E, cork stopper ; 
The weighed F, inverted porcelain dish; G, point of 
quantity of dry sand was first attachment to suction line; H, inner end 
placed 'in each pot and then of suction tube which is protected against 
, , , , . 'the entrance of sand by means of a plug 

thoroughly washed several , 1 . , T A Al , . , ■ : Z 

° - of glass wool. Aote the short length of 

times by covering the surface glasstubing at the bend . This effectively 
with distilled water and draw- prevents the glass wool from being drawn 
the liquid down through the over when suction is applied, 
sand by means of suction ap- 
plied to the outlet tube. Failure of the control cultures to de- 
velop in the sand supplied only with distilled water instead of 
the nutrient solution was sufficient evidence that the washing was ade- 
quate to remove any plant nutrients that might have been pres- 
ent in the unwashed sand. The salts used in making up the cul- 
ture solutions were Baker's " analyzed " monopotassium phosphate 
and calcium nitrate and Merck's "blue label" magnesium sulfate. 
Stock solutions were made up by dissolving gram molecular weights 
of these salts separately in a liter of water. Before making up the 
final nutrient solutions the stock solution was diluted to one fourth 
molecular and stored in flasks which were connected with automatic 



I3O JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



burettes through which the required amount was drawn each time a 
new set of nutrient solutions was to be prepared. 

In each series 36 cultures were employed, each of which received 
at the end of each successive 3-day period a culture solution having 
a total osmotic concentration of 1.75 atmospheres. The solution 
supplied to each particular culture differed, however, from that sup- 
plied to the other cultures in the series, with respect to the propor- 
tions of the three component salts. At the end of each 3-day period 
the pots were weighed and brought back to their original weight by 
the addition of distilled water. The nutrient solutions were then re- 
moved from the pots by suction applied to the outlet tubes and a 
fresh solution supplied through the inverted funnel at the top, thus 
insuring an approximately constant total concentration. 

Discussion of Results, 
arrangement of the cultures. 

For convenience in designating the individual cultures and to give 
clearness to the discussion, the cultures have been grouped in the 
form of an equilateral triangle as has been done for similar series by 
other writers. These groupings are shown in triangle A of figure 23, 
in which the individual cultures are represented by circles. It will 
be seen that the lower row has eight cultures and that as we proceed 
upward each row has one culture less than the one below, the eighth 
row containing but a single individual. The employment of the shaded 
segments to represent the various proportions of the three salts is 
an adaptation of the scheme employed by Harris 5 in his study of 
alkali salts. The unshaded segments in each circle represents the 
number of tenths of the total concentration derived from calcium 
nitrate, the segments marked by small crosses the number of tenths 
derived from monopotassium phosphate, and the stippled segments 
the number of tenths due to magnesium sulfate. Proceeding from 
the base to the apex of the triangle, the rows are numbered from Ri 
to R8, while the individual cultures in each row arc numbered from 
left to right. For example, the second culture from the left in the 
fir-t row is designated R1C2, and similarly the third culture in the 
sixth row i- K6C3. Triangle A of figure 23 shows that all of tjhc 
Solution Supplied to the firsl row of cultures have approximately one 

tenth of their total concentration from monopotassium phosphate 

and those in the < < "ii'l row two tenths, this amount increasing by 

II iTi-. I- S Kffrrt of alkali s;ilts in soils on the Kcrmination and growth 
of cropi. In U. S. Dept Art., Jour. A^r. Research, 5: 12-27. 1915. 



M'CALL & RICHARDS: FOOD REQUIREMENTS OF WHEAT. 131 



increments of one tenth from row to row until the apex is reached. 
As indicated by the shading, the first culture at the left in each row 
has one tenth of its total concentration due to calcium nitrate, this 
partial concentration increasing uniformly by increments of one 
tenth until the opposite side of the triangle is reached. In a similar 



A 



Rl 




MONO-POTASSIUM PHOSPHATE 



C 




MONO— POTASS IUM PHOSPHATE 
SECOND PERIOD 



B 




MONO— POTASS IUM PHOSPHATE 
FIRST PERIOD 




MONO-POTASSIUM PHOSPHATE 
THIRD PERIOD 



Fig. 23. Triangle A shows the salt proportions employed for each culture.; 
and B, C, and D the locations of the best nine and the poorest nine cultures 
for the first, second, and third growth periods, respectively. 



manner the partial concentrations of magnesium sulfate increase 
from right to left in each row. The circle occupying the position 
R2C3 has three tenths of its total area unshaded, two tenths marked 
by crosses and the remaining five tenths stippled, thus indicating that 



132 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



the solution used for this culture had the proportions of three tenths 
calcium nitrate, two tenths monopotassium phosphate, and five tenths 
magnesium sulfate. Throughout the discussion the individual cul- 
tures will be designated by the row number and by the position they 
occupy in the row, using the nomenclature employed by earlier 
writers. 

In order to study better the growth rates, each of the three series 
of 36 cultures have been divided into three groups, (1) a lower 
one fourth comprising the 9 cultures giving the lowest yield of tops, 
(2) an upper one fourth composed of the 9 cultures giving the high- 
est yield, and (3) a medium one half which includes the remaining 
cultures. 

COMPARISON OF THE RESULTS FOR THE THREE GROWTH PERIODS. 

An inspection of the triangular diagrams B, C, and D of figure 23 
will facilitate a general comparison of the three sets of cultures repre- 
senting the three growth periods in the life of the wheat plant. In 
these diagrams the individual cultures are marked with an L if they 
lie in the low-yield group and with an H if they are within the high- 
yield groups. These groups of cultures will be referred to as the 
poorest nine and the best nine in the discussion which follows. 

From triangle B it will be seen that for the early growth period 
the solutions which gave the highest yield of tops are characterized 
by a high calcium nitrate content and a low proportion of mag- 
nesium sulfate, while the lowest yield of tops is associated with low 
calcium nitrate and a high proportion of magnesium sulfate. For 
this period the effect of the monopotassium phosphate appears to 
have been almost entirely overshadowed by the other two component 
salts. A comparison of diagram C, representing the results for the 
second 30-day period, shows a striking similarity in the location of 
the areas of high and low yielding cultures, from which it would 
appear thai mineral food requirements of the wheat plant during the 
lecond growth period were substantially the same as for the first 
30-day period. 

Prom an inspection of triangle D it will be seen that for the third 
and final growth period the solutions which gave the highest yielding 
plant ;nc characterized by a relatively high concentration of calcium 
nitrate and a low proportion not only of magnesium sulfate but also 

of monopota itum phosphate, while the solutions producing low 

yield- are characterized by a high proportion of monopotassium phos- 
phate, without regard to the ratio of the other two salts. These re- 



M'CALL & RICHARDS: FOOD REQUIREMENTS OF WHEAT. 133 

suits suggest a strong possibility that the acidity of the nutrient solu- 
tion may be largely responsible for the location of the area of low 
yielding cultures during the final growth period. The acidity of dif- 
ferent nutrient solutions with special reference to the hydrogen in 
concentration is being made the subject of further investigation. 

The mean molecular proportions of the three component salts and 
the ionic ratios for the culture solutions giving the best and the 
poorest growth of wheat during the different periods are given in 
Table i. 



Table i. — The mean molecular proportions of the three component salts and 
the ionic ratios for the culture solutions giving the best and the poorest 
growth of wheat during the different periods of development. 



Series and growth rank. 


Mean molecular proportions in tenths 
of total concentration. 


Mean cation ratios. 


KH 2 P0 4 . 


Ca(N0 3 ) 2 . 


MgSO«. 


Mg/Ca. 


Mg/K. 


Ca/K. 


Series 1, first ( Best nine 


3-4 


4-7 


1.9 


1.08 


1.50 


I.98 


period. \ Poorest nine . . 


3-8 


2.0 


4.2 


6-55 


2.80 


O.49 


Series 2, second ( Best nine 


3-5 


4-6 


19 


1. 00 


1.48 


I.70 


period \ Poorest nine . . 


2.7 


i-5 


5-8 


8.33 


5-05 


O.71 


Series 3, third f Best nine 


2.0 


5-5 


2-5 


1.32 


2.09 


2.60 


period \ Poorest nine . . 


6.1 


1.8 


2.1 


2 99 


0.52 


0.22 



Table I shows that for the first period the mean molecular propor- 
tions of the solutions producing the best nine are 3.4 parts mono- 
potassium phosphate, 4.7 parts calcium nitrate, and 1.9 parts mag- 
nesium sulfate, and for the poorest nine, 3.8 parts monopotassium 
phosphate, 2.0 parts calcium nitrate, and 4.2 parts magnesium sulfate. 
For the second growth period these proportions remain practically 
the same, but for the final period the proportions are materially 
changed, especially for the low yield group. An inspection of ionic 
ratio values brings out the fact that for the first and second growth 
periods the culture solutions producing the best growth are charac- 
terized by a low ratio of magnesium to calcium, while the solutions 
giving low yields have a high ratio value of magnesium to calcium. 
For the final period this wide difference is absent, the ratio of mag- 
nesium to calcium having increased slightly for the best nine and de- 
creased very materially for the poorest nine. Attention is also called 
to the fact that the mean ratio of magnesium to calcium for the best 
growth for the first and second periods is practically 1 : 1 and for the 
final period 1:1.3. The mean ratio of magnesium to potassium is 
practically the same for the best nine for the first and second growth 
periods, and as in the case of the magnesium-calcium ratio, there is 



134 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

a slight increase for the final period, thus indicating an increase in 
the physiological requirement for magnesium during the late stages 
in the development of the wheat plant. 

Further discussion of these data is being reserved until the conclu- 
sion of similar studies of the mineral food requirements of the soy- 
bean plant and of buckwheat, which are now being conducted by the 
Department of Soils of the Maryland experiment station. 



RELATION OF SIZE OF SAMPLE TO KERNEL-PERCENTAGE 
DETERMINATIONS IN OATS. 1 

R. J. Garber and A. C. Arny. 

In carrying out the rotation investigation and the varietal test 
work at University Farm, the farm crops section of the Minnesota 
station has occasion each year to make a considerable number of de- 
terminations of percentage of kernel in oats (a) of the same variety 
grown in different rotations and (b) of a number of varieties grown 
under similar soil conditions. For oat varieties, these data are of in- 
terest as a basis for comparing true yielding ability and feeding value. 
Inasmuch as hulling oats by hand is rather expensive, it is highly 
desirable to minimize the work so far as is compatible with accuracy. 
The data collected in seeking a solution to this problem are presented 
herein. 

Love (1912) 2 determined for two crops of oats the bushel weight, 
the weight per 100 grains, and the percentage of kernel for each of a 
large number of varieties. The weight per 100 grains was obtained 
by weighing several composite lots of from 50 to 100 grains and 
averaging these for each variety. Kernel percentage determinations 
were made from the weighed samples. Love found a difference of 
nearly 100 percent in the weight pet too grains as determined for the 
different varieties. Comparing the weight per too grains within any 
Variety for the 2 year period, the variation was not great. He found 
.1 ' '<!, ;d'-rablc variation in percentage of kernel for the different 
V&rietil . but no variety showed a wide variation for the 2-year 

1 Pnbli bed with tin- approval of tin- Director as Paper No. 104 of the Jour- 
nal Scries oi the Mmm ot ;i Agricultural Experiment Station, St. Paul, Minn. 

U'(< .''I for puhli<-ation I u << itihcr J>, i<)\7. 

Dattl in parenthesis refer to "Literature cited," p. 142. 



GARBER & ARNY : KERNEL PERCENTAGE IN OATS. 



135 



period. No correlation was found between weight per hundred 
grains and percentage of kernels or between weight per bushel and 
percentage of kernel. 

Surface and Zinn (1916) give data showing no relation between 
weight of a thousand grains and percentage of hull. 

The thirteen varieties studied in this investigation (Table 1) were 



Table i. — Weight of 1,000 grains and mean percentage of kernel in 75 varieties 
of oats with the standard deviation and coefficient of variability in 
twenty 50-grain samples of each variety. 



Variety. 


Group. 


Weight 
of IOOO 
grains. 


Means. 


Standard 
deviations. 


Coefficients 
of 

variability. 


Swedish Select 


Sativa 


29.090 


72. 19===. 14 


.9500=!=. 10 


1-32=1=. 14 


Silver mine 


do 


23.500 


71.73i.13 


.8599=1= .09 


1. 20=l=. 13 




do 


2S4I5 


72.56===. 18 


1. 1795=1=. 13 


1-63=*=. 17 




do 


25.990 


71.92=*= .12 


.7901 =1= .08 


1.10=1= .12 


Improved Ligowa . . 


do 


23-505 


72. 55===. 17 


1.1146=1= .12 


1. 54===. 16 


Sixty Dav 


do 


I8.645 


72.23=±=.I 4 


.9202=1=. 10 


1.27=*= .14 


Iowa No. 103 


do 


I9-525 


72.39=1=. 19 


1. 2851 =1= .14 


1. 78±. 19 


White Tartar 


Sativa orientalis 


26 620 


74.05=1=. 10 


.6601 =1= .07 


.89=1=. 09 


Black Tartarian. . . . 


do 


2O.41O 


69-75^-15 


.9641=1=. 10 


i-38=i=. 15 




do 


45-455 


62.16===. 18 


1.2248=1= .13 


1.97=1= .21 


Garton No. 748 .... 


do 


21.220 


76.70=*=. 12 


.7924=1=. 08 


1.03=1=. 11 


Red Rustproof 


Sterilis 


33.220 


72.80=1= .11 


• 7055 =1= -08 


.97=1=. 10 


Burt 


do 


24.095 


73.26=*=. 13 


.8675=1= .09 


I.i8±.i3 



grown on University Farm in 191 7 and all but the five last named, 
which grew on the plots devoted to classification work, appeared in 
the regular varietal tests. The oats taken from these two sources 
were thrashed similarly, so their respective percentages of kernel are 
directly comparable. It was thought advisable to select varieties dif- 
fering as widely as possible in size and shape of grain, in percentage 
of kernel, and in other characteristics. By so doing, a range of 62.16 
to 76.70 in percentage of kernel was obtained, which was greater than 
would ordinarily be found in any one variety over a period of years 
and consequently obviated the necessity of using the crop from more 
than one year. 

The method employed in making the kernel determinations was 
as follow : A composite sample of a pound or more was made up for 
each variety by taking portions from various places within the bags 
containing the bulk oats. These composite samples were taken to the 
laboratory and each thoroughly mixed. Then, as desired, a com- 
posite sample was poured into a conical pile, from one side of which 
the samples used in determining the kernel percentages were taken. 
The grains were counted out just as they came and, with the excep- 
tion of broken or diseased ones, which were rejected, no selection 



I36 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



whatever was made. The twenty 50-kernel samples of each variety 
were collected at one time and numbered consecutively from 1 to 20. 
Number rather than weight of grains was chosen as a basis for 
sampling because the former was more convenient. The average 
weight of any sample used may readily be ascertained from the 
weight of 1,000 grains for the same variety, as given in Table 1. 
Table 2 shows the frequency distributions of kernel percentages in 
the different varieties. 



Table 2. — Frequency distributions of kernel percentages of 13 varieties of oats, 
as shozen by determinations of twenty 50-grain samples of each variety. 



Classes. 


Swe 
dish 
Select. 


Silver- 
mine. 


Vic- 
tory. 


Ban- 
ner. 


Im 
Li- 
go wo. 


Sixty- 
Day. 


Iowa 
No. 
103. 


White 
Tar- 
tar. 


Black 
Tar- 
tarian. 


Storm 
King. 


v_j3.rton 
No. 
748. 


Red 
Rust- 
proof. 


Burt. 


60.25 




















3 








60.75 




















I 








61.25 




















2 








61.75 




















3 








62.25 




















2 








62.75 




















2 








63-25 




















3 








63-75 




















3 








64.25 




















1 








* * * 




























67-75 


















I 










68.25 


















I 










68.75 


















2 










69.25 














I 




3 










69- 7 5 




I 


I 












7 










70.25 




I 




I 


2 




I 




1 










7«-75 


2 


2 


I 


I 


I 


2 


I 




3 










71-25 


2 


2 




4 




4 






1 






I 


I 


71-75 


2 


7 


3 


6 


3 


1 


3 




1 






2 


I 


72.25 


3 


5 


3 


2 


I 


5 


2 










3 


2 


72 75 


5 




6 


4 


6 


5 


4 


I 








5 


4 


73-25 


4 




3 


2 


2 




4 


3 








5 


2 


73-75 


X 


I 


1 




4 


3 


1 


4 








4 


5 


74-25 














2 


7 










5 


74 75 






1 










2 












75-25 






1 










3 






2 






75-75 




























76.25 






















4 






76.75 






















5 






77-25 






















4 






77-75 






















4 







Doubl'- oat--, commonly so called where the lemma of the primary 
grain complete)} Or almost completely envelops the secondary grain, 
were counted as one. This accounts for the seemingly high weight 
of r ,000 grains of the Storm King variety. 

The Weighing 01 the ample- before hulling and of the kernels 
afterward were made in grams to three decimal places on a Trocm- 



GARBER & ARNY: KERNEL PERCENTAGE IN OATS. 



137 



ner enclosed balance. The kernels were carefully removed by hand. 3 
In order to insure more accuracy only one person worked on the 
seeds of each variety. 

After the data were collected and the various kernel percentages 
calculated, tables of the same form as Table 3, which is presented as 
a representative one, were prepared for each variety. All the varieties 
were handled exactly alike, so an explanation of one table will suffice. 
The column headed " Number of samples " gives the number of fre- 
quencies involved in determining the statistical constants for the 
various replications. Here an explanation in regard to how the repli- 
cations were made is necessary. As has been stated, the 20 50-ker- 
nel samples were numbered from 1 to 20. To secure 1 replica- 
tion the mean percentages were calculated for samples 1 and 11, 2 
and 12, 3 and 13, etc., until the entire 20 were used to make up the 
resultant 10 mean percentages. For 2 replications, samples 1, 7 and 
13; 2, 8 and 14, etc.; for 3 replications, samples 1, 6, 11 and 16; 2, 7, 
12 and 17, etc. ; for 4 replications, samples 1, 5, 9, 13 and 17, etc. ; for 

5 replications, samples 1, 4, 7, 10, 13 and 16, etc.; for 9 replications, 
samples 1, 3, 5, 7, 9, 11, 13, 15, 17 and 19, etc., were similarly com- 
bined and the means determined. Each mean so obtained represents 
one of the percentage frequencies for that particular replication. 
Thus, where one replication is made, 10 percentage frequencies each 
representing 100 oat grains result ; where two replications are made, 

6 percentage frequencies each representing 150 grains result; and 
so on for the various replications. The first horizontal column of 
Table 3 gives the constants where single 50-grain samples only were 



Table 3. — Variability of percentage of kernel in Swedish Select oats. 



Replications. 


No. of kernels 
in each sample. 


No. of samples. 


Means. 


Standard devi- 
ations. 


Coefficients of 
variability. 


None 


So 


20 


72.IQ±.I4 


.9500=1= .10 


1.32^.14 


One 


100 


10 


72.IQ± .IO 


.4805 =±=.07 


.67±.io 


Two 


150 


6 


72.25*. II 


.3942 ±.08 


•55 ± -n 


Three 


200 


5 


72.IQ± .09 


.2946='= .06 


.41 ± .09 




250 


• 4 


72.19=*= .09 


.2800=*= .07 


•39±.09 


Five 


300 


3 


72.25=t .08 


.2086 =±=.06 


.29± .08 


Nine 


500 


2 


72.20=*= .04 


.0852 ±.03 


.12=*= .04 



considered. Table 1 presents this data for the several varieties 
studied, together with the weight of 1,000 grains of each variety. 

Considering the standard deviations as given in Table 3, it is ap- 
parent that a 50-grain sample with a constant of .9500 zb .10 gave a 

3 The hulling was done by two student assistants, B. A. Holt and G. R. Ko- 
katnur, and the senior author. 



I38 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



less accurate determination than a 100-grain sample with a constant 
of .4805 ± .07. The standard deviation of the former is almost 
double that of the latter. Between 100-grain and 150-grain samples 
there is much less difference, but still some decrease in the standard 
deviation. Relatively the same decrease is shown in going from the 
150-grain to the 200-grain samples. Increasing the sample to 250 
grains gave a standard deviation of .2800 ± .07, which is not a marked 
reduction. However, the use of 300-grain and 500-grain samples 
gave standard deviations of .2086 ± .06 and .0852 ± .03 respectively, 
which are more decided reductions. The difference between this con- 
stant, where one replication representing 100-grain samples and nine 
replications representing 500-grain samples were used, is .3953 ± .08, 
which is more than four times its probable error and consequently 
has some statistical significance. The foregoing observations are 
likewise brought out by considering the coefficients of variability. In 
general, similar tables worked up for each of the other twelve varieties 
gave results concordant with those of Table 1, but there was not 
always the same comparatively uniform diminishment in standard 
deviations and coefficients of variability. This is well shown in 
Table 4. 

It will be noted from Table 3 that the means of the different 
replications vary but slightly, the actual difference between the two 
extremes being but 0.06 percent. This, in general, held true for all 
the varieties. In most cases the difference between the extremes was 
less than 0.05 percent and in no instance was it greater than 0.12 per- 
cent. Uniformity of the means within each variety establishes a close 
correspondence between standard deviations and the related coeffi- 
cients of variability, except that the latter are numerically greater. 
The relatively high value of the means compared with their absolute 
differences within any one variety also serve to make the two indices 
of variation for any one replication closely reciprocal. In view of 
these facts, Table 4 includes only the coefficients of variability rather 
than both the coefficients and the standard deviations, which would 
be largely a duplication. 

Tabic l give* .'ill the coefficients Of variability for the twenty 50- 
graill MUnples and the different replications for each of the thirteen 

varieties, together with the statistical constants of these coefficients. 
Consider the column headed single 50-graifl samples. Here the co- 
effii ient of variability of each of the 50-graif] samples is given oppo- 
site the variety from which the determination was made. It is in- 
teresting to QOte ill this connection that in no case docs the coefficient 



GARBER & ARNY: KERNEL PERCENTAGE IN OATS. I 



-H-H-H-H-H-H-H-W-H-H-H-H-H 

N n MnO 00 O rOO MOO m xn 
MOioOroOrOMrotv-MNO 



-H-H-H-H-H-W-H-H-H-H-H-H-H 

O 't 10 ro "3- n (N ro 10 O 10 



fl-H-H-H-H-H-H-tl-H-H-tl-H-ll 

0\ roO h oo K m ro O O m i> 0\ 



-H-H-H-H-H-H-H-H-H-H-H-+1-H 



m 100 ro O -rj- rj- fM moo o 



+I-H-H-H-H-H-H-H-H-H+I-H-H 



-H-H-H-H-H-H-fl-H-H-H-H-H-fl 
r^oo mo a 10 no r^r-oo 
o o <n t->o N <N lO^tM ^-lOOC 



rtroNMO toa^H m O f 

MMHMMWMOlHNMMH 

-H* i ^' -h' -h' ^' 4l' M -H -H -H* -H* -H 



3 Sfoo 



T3 (Li 

CO CO 



£ -a £ 




■H -H 

oi O io 









ro oo 


o 


°. 06 


-h" 






o 















fO o 

^ ^ -H 



-H -H -H 



X rt ° 

03 C 03 

o3 00 

^ CO <S 



140 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



exceed 2 percent. The mean of the coefficients appears near the 
bottom of the column. The standard deviation and the coefficient of 
variability immediately following were obtained in the usual way. 
In other words, each set of statistical constants near the bottom of 
Table 4 was computed from thirteen variants each of which is an 
index of variation for a definite number of samples composed of a 
definite number of grains. 

Considering the table from the standpoint of varieties, it is found 
that they differ widely in degree of variability. For instance, the 
variety of Storm King, with its comparatively high percentage of 
hull, has a coefficient of variability of 1.97 db .21, which is about twice 
as much as that for Red Rustproof, .97 ± .10. This relationship is 
maintained fairly consistently throughout the various replications. 
By comparing these two extremes, it becomes evident that to secure 
the same accuracy in determining the percentage of kernel, Storm 
King requires a larger sample than Red Rustproof. Some varieties, 
as Black Tartarian, show considerable fluctuation and lack of con- 
sistency for the various replications. This undoubtedly is due to the 
somewhat small number of variants involved. In general, however, 
the varieties tend to corroborate each other. 

Comparing the coefficients of variability in the first and second 
columns, it is shown at a glance that one replication representing 
100-grain samples give much less variability than single 50-grain 
samples. In nine out of thirteen cases the coefficients for one replica- 
tion is greater than for two replications, as is shown by com- 
paring columns two and three. Two replications except in two in- 
stances give higher coefficients than three replications. Between 200- 
grain and 250-grain samples represented by 3 and 4 replications, 
respectively, there is little difference in the coefficients. On the aver- 
age. 500 grains constitute a somewhat more representative sample 
for the determination of percentage of kernel than 300 grains. 

Examine nexl the statistical constants of the coefficients of variabil- 
ity. The mean of each column is not presented as an expression of 
variability for the thirteen varieties considered as one group of 
variates, bul rather as the average of the coefficients of variability for 

th<- thirteen varieties, each of which constitutes a group of variatcs. 
The highest mean is found for the single 50-grain samples and the 
U)W( t for the gOO-graifl lamplcs. The former also has the highest 
standard deviation, bul the latter has the highest coefficient of va- 
riability. Where two replications arc made, the standard deviation of 
oefnYirnl . 1 1 7 ♦ .02, is the lowest. In comparing the means of 
tin r offfirimt foi columns 1 and 2, again a decided advantage is 



GARBER & ARNYI KERNEL PERCENTAGE IN OATS. 



I 4 I 



found in favor of using- one replication rather than single 50-grain 
. samples. While the mean of the coefficients in column 3, .67 ± .02, 
is somewhat lower than that of column 2, .81 ± .05, the difference, 
.14 ± .05, is not quite three times its probable error. The difference 
between the means of the coefficients in columns three and four is 
.17 ± .04. Three replications or 200-grain samples diminish the 
mean of the coefficients for one replication or 100-grain samples 
by .31 db .06, which is five times its probable error and therefore sig- 
nificant. The comparatively low variability of the 200-grain samples 
is another point in their favor. The difference between the means, 
.01 ± .05, of three and four replications certainly has no statistical 
value. The differences between the means of four and five replica- 
tions and of five and nine replications are .12 ± .06 and .16 ± .05, which 
are two and slightly more than three times their respective probable 
errors. Between the means of the coefficients of single 50-grain 
samples and three replications the difference is .83 ± .07, while be- 
tween the three and nine replications the difference is only .29 ± .04. 
In other words, 150 grains added to a 50-grain sample reduces the 
average variability almost three times as much as 300 grains added 
to a 200-grain sample. Although it is evident from the above data 
that the 500-grain sample is the most accurate, it is for the individual 
to determine whether for his purpose it is sufficiently more accurate 
than the 200-grain sample to justify the additional work incidental 
to hulling the larger quantity of oats. 

In order to examine still further what constituted a desirable 
sample of oats for determining the percentage of kernel, 100 samples 
of 50 grains each of Improved Ligowa, which is of a Swedsih Select 
type, were counted out, the kernel percentage determinations made, 
and the replications followed in the same systematic way as in the 
several varieties. The results are tabulated in Table 5. The fre- 



Table 5. — Variability in percentage of kernel of Ligowa oats. 



Replications. 


No. of kernels 
in each sample. 


No. of samples. 


Means. 


Standard 
deviations. 


Coefficients 
f variability. 


None 


50 


IOO 


72 71 ± .07 


I.0638 ± .05 


I.46 ± .07 


One 


IOO 




72.71 ± 08 


.7915 ± .05 


I.09 ± .07 


Two 


150 


33 


72.70 ± .06 


.5161 =*= .04 


.71 ± .06 


Three 


200 


25 


72.71 ± .07 


.5430 i .05 


• 75 =*= -07 


Four 


250 


20 


72.71 ± .08 


.5113 ± .05 


.70 =*= .07 


Five 


300 


16 


72.68 ± .06 


.3545 * -04 


.49 ± .06 


Nine 


500 


10 


72.71 ± .08 


.3774 ± .06 


.52 ± .08 



quency distribution of these 100 samples was as follows: 70.25, 5; 
70.75, 5.; 71.25, 2; 71.75, 15; 72.25, 9; 72.75, 18; 73.25, 18; 73.75, 19; 
74-25, 7; and 74.75, 2. 



142 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



One replication reduces the coefficient of variability of single 50- 
grain samples by .37 ±. 10 and two replications further reduce the 
coefficient by .38 ± .09, thus making a difference between 50-grain 
and 1 50-grain samples of .751+1.09, which is approximately eight 
times its probable error and, therefore, of statistical value. Con- 
sidering replications two, three, and four, no difference of statistical 
significance appears ; likewise, between five and nine replications. The 
difference between four and five replications is only .21 ± .10, while 
between five and nine replications, it is negligible. 

In view of the foregoing facts a 1 50-grain sample of Improved 
Ligowa oats taken as described gave as statistically representative a 
sample as a 200-grain or a 500-grain sample. These results with the 
large number of grains of a single variety are in substantial agree- 
ment with the interpretation of the data as given for the thirteen 
varieties. 

Conclusions. 

From the data presented in this study, the following conclusions 
may be drawn : 

1. In general, a 200-grain sample taken as has been described 
gives a sufficiently accurate determination of percentage of kernel in 
oats for all ordinary purposes. The weight of this size of sample 
varies from 3 grams in early to 5 grams in midseason and late 
varieties in Minnesota. 

2. Where more than ordinary accuracy is necessary, the sample 
should be increased to at least 300 grains, and with some varieties 
still larger samples are desirable. 

LITERATURE ClTED. 

LOVF. H. H 

1912. Oats for New York. N. Y. Cornell Univ. Agr. Expt. Sta. Bui. 343, 
p. 407-416. 
Surface, Frank M., and Zinn, Jacob. 

1916. Studies on oats breeding. Maine Agr. Exp. Sta. Bui. 250, p. 140-141. 



AGRONOMIC AFFAIRS. 



H3 



AGRONOMIC AFFAIRS. 

MEMBERSHIP CHANGES. 

The membership reported in the January issue was 656. Since that 
time 16 new members have been added, 3 have resigned, 1 has died, 
and 30 have lapsed for nonpayment of dues of 1916, a net loss of 18 
and a present membership of 638. The names and addresses of the 
new members, names of those lapsed, deceased, and resigned, and such 
changes of address as have been noted since the last issue, are as 
follows : 

New Members. 

Anderson, Arthur, Univ. Farm, Dept. of Agronomy, Lincoln, Neb. 

Barbee, O. E., Cliff House, Pullman, Wash. 

Berry, Roger E., 404 Knoblock St., Stillwater, Okla. 

Gordon, Thomas B., Mass. Agric. College, Amherst, Mass. 

Haseltine, L. E., 2301 Durant Ave., Berkeley, Calif. 

Hatcher, Otto, 318 West St., Stillwater, Okla. 

Hildebrand, E. B., 112 N. Main St., Stillwater, Okla. 

Johnson, D. R., 220 Knoblock St., Stillwater, Okla. 

Johnson, George F., 256 W. Woodruff Ave., Columbus, O. 

Kearney, T. H., Bureau of Plant Industry, Washington, D. C. 

McDowell, C. H., Supt. Substation No. 6, Denton, Texas. 

McGuffey, C. Carl, McGuffey, Ohio. 

Thompson, R. S., Highland, Calif. 

Turner, A. F., K. S. A. C, Manhattan, Kans. 

Wells, W. G., Bur. Plant Ind., U. S. D. A., Washington, D. C. 

Wunsch, W. A., Acting Co. Agt, Newton, Kans. 

Zinn, Jacob, Agric. Experiment Station, Orono, Me. 



Edwards, R. W. 



Members Resigned. 
Kyle, C. H., 



WORTHEN, E. L. 



L. J. Briggs. 
I. N. Chapman. 
Ernest W. Curtis. 
N. C. Donaldson. 
F. V. Emerson. 
Raymond C. Gauch. 
E. W. Hall. 
J. D. Harper. 
S. H. Hastings. 
D. S. Jennings. 



Members Lapsed. 

Grover Kinzy. 
W. A. Lintner. 
R. E. Lofinck. 
Jas. McAdams. 
G. W. Morgan. 
D. S. Myer. 
A. J. Ogaard. 

W. M. OSBORN. 

A. W. Palm. 
J. W. Paxman. 



Martin Reinholt. 
R. R. Reppert. 
A. M. Richardson. 
J. C. Robert. 
C. H. Ruzicka. 
E. B. Watson. 
Isaac Weisbeim 
W. A. Wheeler. 
Casper A. Wood. 
Harry P. Young. 



144 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Members Deceased. 
Malon Yoder. 

Change of Address. 

P. V. Cardox, Experiment Farm, Moccasin, Montana. 

Foersterlixg. H., care The Arbor Farm, Jamesburg, N. J. 

Foord. J. A., 54 Lincoln Ave., Amherst, Mass. 

Jexsex. L. X.. Box 1214. Amarillo, Texas. 

Mackie, W. W., 121 Hilgard Hall, Berkeley, Cal. 

Westover. H. L., Bard, California. 

Wheeler. H. C, Agric. Bldg., Urbana, 111. 

Woodard. John, 381 Paisley Road, Guelph, Ont., Canada. 

NOTES AND NEWS. 

\. R. Mann, who has been acting dean of the college of agriculture 
and director of the experiment station of Cornell University, has been 
made dean and director. 

Walter Packard, formerly in charge of the Imperial Valley Experi- 
ment Farm at El Centro, Cal., is now chief of farm advisors for the 
southern half of California, with headquarters at Berkeley. 

J. T. Parsons has been appointed soil technologist and O. I. Snapp 
has been made assistant in soils at the Ohio station. 

Gordon W. Randlett, for the past two years director of extension 
in South Dakota, has been elected to a similar position in North 
Dakota. 

Bernard F. Sheehan, formerly of the Iowa station, has been ap- 
pointed instructor in farm crops at the Oregon college and station. 

L. Van Es, for many years veterinarian of the North Dakota 
Station, has been elected director of that station. 

< \V. Warburtcn, a member of the Committee on Seed Stocks 
of the I*. S. Department of Agriculture, is now engaged in the pur- 
hasc of seed grain for resale to farmers at cost in the sections of 
North Dakota and Montana where crops failed last year. This work 
is tinder the provision- <>i the emergency food production bill. Mr. 

Warburton'i headquarters are in Minneapolis while engaged on this 
work. 

\V. Iv. Ward, in tractor in agronomy, and Horace J. Young, as- 
- i~taiit professor Of agronomy, both of the University of Nebraska, 
■ ergenci di tricl demonstration leaders in that state. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. io. April, 1918. No. 4. 

METHODS USED AND RESULTS OBTAINED IN CEREAL 
INVESTIGATIONS AT THE CORNELL STATION. 1 

H. H. Love and W. T. Craig. 

Introduction. 

During the past few years considerable interest has been mani- 
fested in the methods used in the various experiments of crop re- 
search, and considerable change in the handling of experiments in all 
agronomic investigations has resulted. This has been especially true 
with regard to the small grains. When, in addition to testing a few 
of the better-known varieties, selection and breeding work was un- 
dertaken, it was important to be able to handle a large number of sorts. 
This demanded a change in methods so that fair comparative results 
could be obtained without unnecessarily large acreage being devoted 
to any certain crop. The tendency has been to reduce the size of plat 
more and more until finally in a number of places the rod-row system 
has been adopted. So far as is known to the authors Mr. J. B. Nor- 
ton, of the United States Department of Agriculture, was the first 
to put this method in general use. 

Montgomery 2 has discussed this method in relation to plat trials. 
Since the authors in their work at the Cornell University Agricul- 
tural Experiment Station 3 have for a number of years been using 

1 Paper No. 64, Department of Plant Breeding, Cornell University, Ithaca, 
New York. Presented by the senior author, with illustrations, at the tenth 
annual meeting of the American Society of Agronomy, Washington, D. C, 
November 13, 1917. 

2 Montgomery, E. G. Experiments in wheat breeding ; experimental error 
in the nursery and variation in nitrogen and yield. U. S. Dept. Agr., Bur. 
Plant Indus. Bui. 269, 1913. 

3 This work is in cooperation with the Bureau of Plant Industry, United 
States Department of Agriculture. 

145 



I46 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



this method rather extensively in connection with cereal breeding 
work it seems worth while to present some description regarding its 
details. In addition, the methods used in selection and hybridization 
will be discussed, with a brief statement of the results obtained. 

Methods Used in Making and Testing Selections. 

Two methods are used in making selections with which to begin 
comparative trials. The first is to travel through the State making 
selections from various fields. In making these selections the dif- 
ferent types present in the field are collected, as well as the main 
type represented. Care is taken to select from that part of the field 
which represents average conditions as nearly as can be ascertained. 
Heads are selected rather than plants, owing to the difficulty of sep- 
arating individual plants with certainty. In this sort of selection it 
is important to make collections from as many different fields as pos- 
sible and to cover a wide area, so that various growth conditions may 
be encountered. 

The second method followed is to make plantings of a selected 
variety or varieties from which selections are to be made. Indi- 
vidual kernels are planted 1 foot apart each way in ground previously 
prepared and marked off. In this way it is possible to study the en- 
tire plant and make comparisons as to stooling habits, stiffness of 
straw, and the like. An objection to this method is that it is not pos- 
sible to make selections on such a large scale as when they are made 
from the field. That is, from the field the selections represent the 
best from a vast number while from the hill method they represent 
the best from only several thousand at most. This objection is partly 
offset by the fact that a more detailed study can be made of the indi- 
viduals selected and that a larger quantity of seed is obtained than is 
possible from a single head. In either case it is very important to 
Mart with a very large number of heads or plants. 

After these heads are selected they are then thrashed and prepared 
for planting. Few notes arc taken on the heads as the more detailed 
Btlldy of the new sorts is made later on in the trials. For thrashing 
tl ese heads a small head-thrasher has been constructed by Mr. H. W. 
Teeter, of the Department of Plant Breeding. No screen or fan is 
attached to tlii machine, as it is used also for a great amount of sta- 
• tical work where all kernels, even the light ones, arc to be saved, 
mall electric motor furnishes the power. Very rapid work can 

be done with tlm machine, which is operated by one man. A hun- 
dred or more heads can be thrashed and cleaned per hour. 



LOVE & CRAIG : CEREAL BREEDING METHODS. 



H7 



When heads are selected the seed is sown in head rows the first 
year. In order to have a uniform seeding the seeds are counted out 
so that the same number may be sown in each row. The seeds are 
placed in an envelope and this envelope numbered with the proper 
row number. Sometimes it is necessary to sow in two lots, as it is 
not always possible to obtain the same number of seeds from each 
of the heads chosen. For example, it may be necessary to plant a 
series where 30 seeds per row and another where 25 seeds per row 
are used. 

For wheat these head-rows are now 2.5 feet long and for oats 5 
feet long. Usually 25 or 30 seeds are used in each head row for 
wheat and 40 to 60 for oats. As stated above, the number of seeds 
used depends on conditions affecting the size of head selected. 

These seeds are scattered in the drills thinly and are not spaced, 
as experience has shown that for such work careful spacing is not 
necessary. It is more important to have a very large number of 
selections under test and depend on them to furnish something worth 
while than to handle a few selections with greater attention to detail. 
For example, this year we have under test 2,200 new selections of 
wheat and two years ago we had 1,600. It is our plan to start a new 
selection series every two years. This works out to better advantage 
than starting a series every year, as it gives time for the elimination 
of the poorer sorts. The number of head rows of wheat that can be 
sown in a day by five men is about 2,000 to 2,500, while for oats it is 
less, as the row is longer and more seeds are to be dropped. 

If plants rather than heads have been used as a basis for selection 
then more seed is available and a definite quantity of seed may be 
weighed out for each plant row. Weighing may be objected to since 
the number of seeds from a large-seeded sort will be less than from 
a small-seeded one. It has been shown, however, that within a rather 
wide range the difference in number of seeds will not affect the final 
results. With plants it will be possible also to have longer rows 
than with heads. 

It may be worth while to mention how these drill rows are pre- 
pared for sowing. The land to be used is put in as good tilth as 
possible. Then a sled marker with runners a foot apart is used to 
mark off the field. This is then followed with a single-shovel plow 
to open the drills. When the ground is in good condition a little 
Planet, Jr., plow with disk attachment is used for covering ; other- 
wise, hoes are used. The marker and drill are started a little while 
before sowing, marking and opening the drills across the plat of 



I48 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

ground to be used. This makes for more rapid work, as the rows are 
all laid out before any stakes or lines are placed. The rows and 
paths between the series can then be measured off and lines run to 
mark the end limits for a large number of rows at once. This plan 
is used whether head or rod rows are to be sown. 

While the plants in the head rows are growing and developing gen- 
eral notes are taken which with the yield of grain help to determine 
the worth of a particular selection. At harvest time the rows are 
studied, usually by two persons, and those rows that are apparently 
superior are harvested and tagged. No selection of plants from these 
rows is made, as it is not possible to change the type by reselecting 
plants from these rows which, barring accidental crossing, are already 
pure lines. Some natural crossing occurs with wheat; these hybrid 
rows are usually discarded. No natural oat hybrid has yet been 
found. A large number of the head rows are left in the field because 
they show clearly that they are not worth continuing. It may be 
argued that one year's test is not sufficient to determine the value 
of a selection. This may be partially true, yet if these selections have 
been made from varieties adapted to the locality it would be expected 
that the better selections would show up well the first year. Then, 
too. the selection of these head rows is not made so closely but that 
practically all of the better ones are retained. About three fourths 
of the rows are usually discarded the first year. We feel that by 
discarding in this manner and bringing in new selections every second 
year as indicated greater progress will be made than by keeping a 
large number of strains of doubtful value. 

These head rows are then thrashed and the grain weighed and re- 
cleaned for sowing the second year. After thrashing a few of the 
poorest may be discarded, although as a rule those that have passed 
the field inspection are retained for at least one more year. These 
selections are sown in triplicate rod rows the second year provided 
enough seed; if not, they are sown in duplicate rod rows. 
I he amount of seed to be sown per row is calculated from the 
usual rate of seeding for any particular grain. The seed is weighed 
and put into envelopes and the envelopes numbered with the proper 
according to tin plan of planting. After all of the en- 
velope are numbered these numbers are checked with the plan of 
planting to be lure all are in their proper order. A check row is 
OWH ' ■' '■■ tenth row. Seed for this check is taken from a standard 

sort or an improved strain. The cheek rows are marked with num- 
bered Stakes which serve as an aid in taking notes and harvesting. 



LOVE & CRAIG : CEREAL BREEDING METHODS. 1 49 

In this way any variations in soil differences may be determined. 
The seed is sown by hand from the envelopes and it is possible for 
five men to sow 1,000 to 1,200 rows a day. It is found that with a 
little experience in sowing the seed can be distributed just as evenly 
as with a drill. 

The plan of planting is arranged so that no two rows sown from 
the same sort come together. For example, if there are 144 sorts to 
be tested these together with the check rows will make 160 rows. 
This series of 160 rows is sown and the series is then repeated; that 
is, row 161 is the same as row 1, 162 the same as 2, and so on for the 
entire series. In order to prevent any effect which may be caused 
by two unlike sorts growing together the different strains are ar- 
ranged according to earliness and other characters so as to reduce 
this source of error to a minimum. Some of the rod rows of wheat 
are shown in Plate 4, figure 1. 

The length of these so-called rod rows varies, depending on the 
kind of crop being handled. All of the weights of grain are taken 
in grams and the yields per row estimated in bushels per acre. In 
order to make these calculations as simple as possible the length of 
the row is so changed that some simple factor is used to convert 
grams per row into bushels per acre. The oat rows are 15 feet long 
and the grams per row are multiplied by 0.2 to obtain bushels per 
acre. For wheat the length is 16 feet and for barley 20 feet, the con- 
version factor in each case being o.i. 4 

In every case it is desirable to sow longer rows than are harvested. 
With oats and barley these rows are sown 1 foot longer, or 16 and 
21 feet, respectively. For wheat, which must stand through the 
winter, the row is 2 feet longer or 18 feet. It is obvious that if the 
end of each row is cut off, more nearly uniform conditions may be 
obtained and the effect of increased nutrition which occurs at the 
ends will not enter into the calculations and modify the results. 

During the growing season any desired notes are taken on the 
rows, such as disease resistance, height, habit of growth, type of 
head, and the like. At harvest time the rows are cut and carefully 
tagged and bound. The bundles are bound near the butt with twine, 
while wired Denison tags which have been previously numbered in 
accordance with the plan of planting are used to tie the bundle near 
the heads. When harvesting rod rows it is better to have three men 
in a team, two to cut and one to tie and label the bundles. A pole 

4 These factors are in reality 0.200066+ an d 0.100033+5 but the remainder 
of the fraction is so small that it does not affect the result when only the first 
significant number is used in each case. 



I 50 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

of the proper length is used to cut the rows by so that the exact 
length is harvested. Three men will cut from 300 to 400 rows a 
day. The bundles are not bagged to prevent mixing but are taken 
immediately into the thrashing shed and hung heads down in the same 
order that they are grown in the field. This method of hanging 
(Plate 4, fig. 2) was designed by the junior author and is so planned 
that ten bundles can hang in a row. These are hung on nails driven 
into strips. The system is flexible so that the different quantities of 
straw can be accommodated. These strips rest on supports which 
are hung from the roof of the building. These supports are 2X6 
inches and one of them is notched at various distances so that the 
strips may be held in an upright position. These notches make it 
possible to shift the strips in accordance with the quantities of straw 
and size of the bundle. This system allows the grain to dry thor- 
oughly without mixing and keeps it in good condition until thrashing. 
"When the grain is cared for in this manner the thrashing of oats and 
barley may be left until the wheat is thrashed and sown again and 
the other summer work finished. 

At thrashing time the three bundles of each soft are gotten out, 
which is easily done by the system of hanging. These bundles are 
thrashed individually, but it is not necessary to clean the machine 
thoroughly after every row as would be the case if the bundles were 
taken out of a pile just as they came. It is only necessary to 
clean the machine thoroughly after each variety, thus facilitating 
thrashing. Before thrashing the total weight of straw and grain is 
taken and after the grain is thrashed and weighed the yield of straw 
may be obtained by taking the difference between total weight and 
weight of grain. 

The thrashing of these rod rows is done by means of a specially 
constructed machine designed by Mr. H. W. Teeter, of the Depart- 
ment of I 'lant Breeding. This machine is so built that the screens 
are in sight and after finishing with one sort if any grain remains 
behind it can be seen and brushed out. The thrashing can best be 
done by four men. One man can follow the plan of planting and get 
out the bundles, one do the feeding, one bag and label the grain, and 
the fourth can look after the machine and take care of the straw, 
[t i ible I ( I thrash about 500 to 600 rows of wheat and 450 to 

'->> row of in a day. 

the differed rows of each sort are thrashed at the same time 
can be kept together and the work thus simplified. After the 
individual Wtightl Of row are obtained the grain from the several 



Journal of the American Society of Agronomy. 



Plate 4. 




LOVE & CRAIG : CEREAL BREEDING METHODS. I 5 I 

rows is brought together, weight per bushel tests are obtained, and 
the seed prepared for sowing another year. These weights per 
bushel are taken by means of an apparatus described in a previous 
article, 5 which gives very accurate results. The variation between 
different determinations of the same variety is very small. This ap- 
paratus is similar in operation to one described more recently by E. 
G. Boerner. 6 

The selections for further testing are then made, basing them on 
yield and other qualifications. In the third year of this test those 
sorts that are continued in the rod-row series are repeated ten times, 
with a check every tenth row the same as the second year (Plate 4, 
fig. 1). The method of handling, so far as sowing, measuring the 
rows, and the like are concerned, is the same as has been described. 

At thrashing time the ten bundles of a sort are brought together. 
These are thrashed by individual rows in order to have the data to 
study the variation of the same sort in different parts of the field. 
If such data are not desired then it is possible to thrash all ten 
bundles together, letting the total weight represent the yield of the 
sort under the conditions tested. This eliminates a great deal of the 
detail work of thrashing, thus making it possible to thrash about 30 
sorts of 10 rows each per hour. 

The different strains are continued in the rod-row series for at 
least three years before any one is eliminated unless it is evident that 
a particular strain is not at all adapted to the given locality. In ad- 
dition to the rod-row series a few of the better selections which give 
evidence of superior merit are grown in increase plats. These in- 
crease plats vary in size from year to year, depending on the number 
of sorts to be grown and the land available. They are so planned, 
however, that they may be run in duplicate or triplicate and therefore 
be of use as variety tests. 

Any new sorts which have given good yields for a number of years 
are then tested further by sending seed to farmers who are willing 
to cooperate to the extent of making comparative trials of the new 
sort with the varieties grown in their locality. 

The value of the rod-row system has been much discussed and 
various objections have been made to its use. The tendency now, 
however, is to give this method its rightful place in agronomic technic. 

5 Love, H. H. Methods of determining weight per bushel. In Jour. Amer. 
Soc. Agron., 7 : 121-128. 1915. 

6 Boerner, E. G. Improved apparatus for determining the test weight of 
grain, with a standard method of making the test. U. S. Dept. Agr. Bui. 472. 
1916. 



152 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



To those who have given careful thought and study to plat methods 
it is evident that by using a rod-row system and replicating a suffi- 
cient number of times the probable error of the average yield is re- 
duced to a very low degree. This is illustrated by some data pre- 
sented here which are taken from some of the rod-row records. 

"Wood and Stratton 7 have said regarding the use of small plats 
that for very small plats of one-one thousandth to one-five thou- 
sandth of an acre the probable error may be reduced by replicating 
the plats systematically. They predicted that the probable error of 
one such plat could be reduced to about 12 percent and that for nine 
such plats to about 4 percent.. 

The rod-row plats are about half way between the two sizes of plat 
mentioned. It is of interest then to present some data on the yields 
of rod rows repeated ten times. These yields, with the means and 
probable errors, are given in Table 1 for some varieties of wheat and 
oats. 

Table i. — The average yield and probable errors and probable error as per- 
centage of the mean of some varieties of wheat and oats tested by 
the rod-row method. 

WHEAT. 







Probable error as 


Variety. 


Mean and prob- 


percentage of 


able error. 


the mean. 




. . 44.7O + I.42 


3.18 


Dawson Golden Chaff 522-68 , , 




2.38 


Gypsy 


. . 41.88+ 1.23 


2.94 




. . 39.I\I + I.06 


2.71 




. . 39-04+ 146 


3-74 


Rural New Yorker No. 57 


•• 37-94+ -97 


2.56 


Fishhcad 


.. 37.17 ± 48 


1.29 






2.69 


OATS. 






Welcome selection 123-5 


. 80.4O + 2.90 


3.6l 




. . 76.64 + 2.33 


2.90 


Silvermine selection 120-9 


• • 75-20 ± 2.19 


2.91 


Sixty Day 593^1 


, 73.60 4- 1.22 


1.66 






327 




. . 72.18 + 2.60 


360 


Burt X Sixty Day 


.. 59.52 + 2.32 


300 


Average 




312 



I bete arc only a few of the many tests, yet as the varieties were 
< 1' < iv<\ al raivloni tin v fairly represent the average condition for 

1 \\'o'»\. 1 V, , and Stratton, F, J M. The interpretation of experimental 

result*, in Jour. A«r. Sci., vol 3, part 4. 1910. 



LOVE & CRAIG : CEREAL BREEDING METHODS. 



153 



wheat and oats. It is seen that the probable error of the mean, ex- 
pressed in percentage, varies between 1.29 and 3.74 for wheat and 
between 1.66 and 3.90 for oats. The average of the probable errors 
expressed in percentage is 2.69 for wheat and 3.12 for oats, which is 
well within the limits predicted for such work by Wood and Stratton. 

The relation between the rod-row yield and plat yield for a few 
varieties grown the same year has been plotted graphically (fig. 24) 
in order to show the relation between the yields of the same strains 
tested under the different conditions. The broken line represents the 
yield of the rod rows of wheat for 1916, while the solid line repre- 
sents the yield of the same strains tested on one two-hundredth-acre 
plats repeated three times. The yields do not follow in exact order 
but in general the relationship holds. This is well illustrated by 
means of the straight line fitted to the plat yields. The equation to 
this line is 3> = 37.253 + .177^*. When one considers the fact that for 
the rod rows and drill plats it is very difficult to have all conditions 
similar the variation between the two methods is not unexpected. 
The relation between the yields of the two methods is practically the 
same as the data published by Montgomery show. 




Fig. 24. Graph showing relation of yield of wheat varieties in 1916 when 
same sorts were grown in rod rows repeated ten times and in two-hundredth- 
acre plats repeated three times. 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



This chart also shows a marked difference between the yields of 
the rod rows and the plat yields. The plat yields are with one ex- 
ception considerably higher than the row yields. The average for 
all the plats is 40 bushels per acre, while for the rows of the same 
varieties it is 31.4 bushels. This overcomes the objection often made 
against the rod-row system that the yields are higher than plat trials. 
To be sure, there is some difference in stand under the two systems. 

As the probable error is reduced so low with rod rows one seems 
justified in using this method, as so many more sorts may be handled 
on a given area. For example, with oats when the rod rows are 
repeated ten times they comprise altogether about one two-hundred- 
ninetieth of an acre which, together with the required paths between 
the series, makes it possible to handle about 242 sorts repeated ten 
times on an acre. On the other hand, with two-hundredth-acre 
plats repeated three times and with 2-foot borders between, it is pos- 
sible to handle only about 37 sorts. With these facts in mind we 
feel that for our purpose the rod-row method is the best. 

Table 2. — Thrcc-ycar average yields in bushels per acre of oat and wheat 
varieties and selections, with the increases obtained. 

OATS. 

3-year average 
yield per acre. Gain. 



Variety and selection numbers. Bushels. Bushtls. 

Canada Cluster 56.9 

Canada Cluster 110-36 65.6 8.7 

Lincoln 53.2 

Lincoln 100-14 57.4 4.2 

Lincoln 109-15 ' 58.4 5.2 

Big Four 52.4 

Big Four 115-27 58.3 5.9 

Big Four 115-40 57.9 5.5 

Clydesdale 50.2 

Clydesdale 114-2 56.7 6.5 

( 1\ desdale 1 14 4 57.1 6.9 

Clydesdale 1 14-14 57.8 7.6 

Clydesdale 1 14-16 58.0 7.8 

WHEAT. 

Klondykc 28.2 

Klondykc 126-26 30.4 2.2 

Klondykc 12^44 31.3 3.1 

Fulcastcr 26.0 

Fulcastcr 123-23 27.9 1.9 

Fulcastcr 123-32 (beardless) 30.2 4.2 

Red Wave 27.7 

\<vi\ Wave 12K 47 3 1 . 1 3.4 



LOVE & CRAIG : CEREAL BREEDING METHODS. J .55 

Some of the results obtained by this method of selection and test- 
ing are shown in Table 2. 

It may be well to summarize the methods used in making selections. 
The heads or plants are selected and tested for one year in head or 
plant rows. The best rows are selected in the field, harvested, 
thrashed and grown the second year in rod rows repeated two or 
three times, depending on the amount of seed available. Only the 
very poorest are eliminated the second year. The rest are continued 
in rod rows repeated ten times for at least three years. The best 
new strains are multiplied and tested in increase plats. The best 
ones are finally distributed to farmers for further comparison. In 
this way a sort is tested for at least six years before it is finally put 
into general use. 

Methods Used in Making and Studying Hybrids. 

In making hybrids of the small grains it has been found best to 
grow the plants to be used for hybridization in pots in the green- 
house. Experience has shown that in most seasons when the plants 
are grown in the field the hot sun dries up a large number of the 
flowers which have been emasculated, thus greatly interfering with 
the work. Another point in favor of growing greenhouse plants is 
that pollen is available for more hours per day and that the pots can 
be moved about, thus bringing those being worked close together. 
The plants, especially in the northern States, develop more slowly in 
the greenhouse, thus giving more time to complete the work. An- 
other very important point is that the work may be done in the late 
winter or early spring before the heavy field work of summer comes. 
Loss from injury by storms is also avoided. 

In crossing oats several spikelets on a head are chosen at a time 
when the anthers are still young and are green in appearance. The 
smaller or upper flower (or flowers) is removed from each spikelet 
and the anthers removed from the remaining flowers by means of a 
small pair of forceps. The spikelets are then tagged with a tag 
which shows the date of emasculation. After a day or two the flowers 
are pollinated. This is done usually by collecting anthers that are 
just ready to burst, placing one in the flower, and carefully replacing 
the palea. Sometimes when plenty of pollen can be obtained the 
pistils are dusted with pollen. At other times anthers which are 
nearly ripe are taken from the plants used for male parents and 
dropped into the flowers at the time of emasculation. All these 
methods have given very good results. 



I 56 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



In all cases a record is kept showing which particular plant of a 
- arioty was used as a male parent. When the hybrid seeds are ripe 
they are collected, as are also seeds from both the male and the 
female parents. These seeds are sown in pots again in the green- 
house. Greater care can be given them there than in the field and 
there is a greater possibility of saving all the plants. Seeds from 
each of the parents are also grown in order to make direct compari- 
sons between the F t plant and the parent types. The seeds from 
the Fj plants as well as the parents are saved and the various charac- 
ters described and noted. 

The Fo and succeeding generations are grown in the field in 5-foot 
rows. The rows are prepared in the same manner as for the rod 
rows. In sowing, the kernels are spaced 2 to 3 inches apart in the 
row. This spacing is not made by exact measurement, as that is un- 
necessary. One row of each of the parents is also sown for com- 
parison. The 5-foot rows with a 2-foot border between the sections 
permit the taking of notes on the plants without interfering with their 
growth. It is also possible to sow more rapidly, as it is possible to 
guage the spaces needed by the number of seeds in hand. A row 
may be confined to a single sort, while if longer rows are used it is 
necessary sometimes to have more than one sort in a row, which 
leads to confusion and possibly to errors. 

At harvest time the plants are pulled if it seems necessary to save 
seed from a large number. When this is not needed the plants are 
pulled and a culm cut from each for further study. By saving only 
a culm it is possible to store a large number of families in a small 
Space. When this latter method is used all of the heads from a 
fairly large number of different types are saved in large envelopes. 
In this way seed is retained for the next planting. Unless one saves 
r rather large number it may be found that not all the types desired 
arc present, in certain wild oat crosses it is necessary to use en- 
velope! for many of the plants to prevent loss by shattering. 

When wheat is to be crossed some of the spikelets are removed 
near the middle of the head, and either the upper or lower flowers 
fttUHCIllated. All but the OUtside flowers of the spikclet arc removed. 
In ome . th< tip of the glumes are cut off and the anthers re- 
movedj While at other times the anthers may he removed without clip- 
ping tlx- flumes. 'Ill, lu-ub are then covered with glassine bags 
for the purpose. Kither way of preparing the head is satis- 
factory, as the data in Table 3 show : 



LOVE & CRAIG : CEREAL BREEDING METHODS. I 5 7 

Table 3. — Results of two methods of emasculation for wheat pollination. 

Number of heads worked 90 

Number of heads setting seed 67 

Percentage of heads setting seed 74-44 

Percentage of heads setting seed when glumes were clipped 74-35 

Percentage of heads setting seed when glumes were not clipped . . 74.51 

Average number of seed set 7.2 

Average number of seed set when clipped 8.5 

Average number of seed set when not clipped 6.2 



From these data it is seen that there is practically no difference as 
to results whether the glumes are clipped or not. This method may 
be objected to because pollen from the non-emasculated flowers of the 
head may adhere to the glassine bags so that when they are removed 
for pollination some of it may fall on the emasculated flowers. Ex- 
perience does not seem to give this objection any weight. Wheat 
hybrids in general are handled in the same way. 

Most of the notes on these hybrids are taken in the laboratory 
rather than in the field, though certain notes which can be more 
readily taken in the field are secured. It is difficult to handle as 
large a number as we now have and take many notes in the field, 
particularly as the various sorts soon begin to weather badly and 
therefore are much discolored. Again, certain crosses of Avena 
species shatter badly unless the plants are harvested rather early. 

In addition to the studies in inheritance, any hybrids that give 
promise of commercial value after becoming fixed as to their char- 
acters are placed in the rod-row series for testing. 

Space does not permit a full account of the results obtained, yet 
it is well to mention a few. With oats, color studies have been made 
in various crosses. In Avena fatua\A. sativa (variety Sixty-Day) 
crosses it is found that the yellow color of the Sixty-Day apparently 
inhibits awn production as well as the production of pubescence. In 
a cross between two black oats classed as the same type, non-black 
oats appear in the second generation in the ratio of 15 black to 1 non- 
black. Crosses between hulled and naked oats show that these char- 
acters behave as a simple monohybrid, but that there is apparently 
some modifying factor present which affects the amount of hull in 
the heterozygous forms. 

With wheat the red color of kernel is found to be represented by 
one, two, and three factors in inheritance. This is in accordance with 
the results of other workers. Many crosses between the different 
species are being studied. Two fertile wheat-rye hybrids have been 
found, one of which has been carried to the fourth generation. 



I58 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



A SIMPLE METHOD OF DEMONSTRATING THE ACTION OF 

LIME IN SOILS. 1 

Paul Emerson. 

Every demonstrator or lecturer realizes the fact that the more 
vividly he paints his. word pictures the more lasting is the impres- 
sion made on his audience. If the lecture is illustrated it is better 
understood, even with a poor lecturer, than if not, and the closer the 
illustrations are to the point the better. 

While demonstrating methods for testing soils and determining 
their lime requirements at a large State fair recently, the writer was 
asked this very pointed question : " How can you prove to me that 
limestone has a beneficial action in the soil by neutralizing the acid- 
ity?" 

The questioner was of the type that demands to be shown rather 
than told, and his question was answered in the following manner. 
A few drops of sodium alazarin sulfonate (alazarin red) indicator 
was put into about 100 c.c. neutral water in a flask. The farmer 
was asked to note the resulting brown color and also any other change 
of color that might subsequently appear. Approximately 2 grams of 
ground limestone were then dropped into the water and the flask 
thoroughly shaken. Naturally the characteristic alkali color appeared. 
This state was assumed to be that of a soil in good tilth. But prac- 
tically all farming operations tend toward an acid reaction, so a few 
drops of dilute hydrochloric was added. The farmer was quick to 
note the solution's change from red to yellow, but when the flask was 
agitated a few times and the alkali color returned, he was amazed. 
W hen the phenomenon was explained to him in a manner that applied 
to the soil, he was apparently convinced that the beneficial action of 
■ ' ' 1 ' 1 ' ■ in the soil continued as long as there was any present. 
Thia demonstration proved to be very popular and absolutely settled 
umber of arguments. It is hoped that lecturers on limes and lim- 
ing will be able to make me of the method, as the materials required 
take up little km .in and are found in practically all laboratories. 

I Contribution from the Maryland State Agricultural Experiment Station, 
CoIIckc Park. lid. Received for publication November 13, 1917. 



m'kee: glandular pubescence in medicago. 159 



GLANDULAR PUBESCENCE IN VARIOUS MEDICAGO SPECIES. 1 

Roland McKee. 

In a study of the occurrence of glandular pubescence in the various 
species of Medicago, an attempt was made to determine to what ex- 
tent this character varies with environmental conditions. It has been 
observed that glandular pubescence is much more strongly developed 
in some species than in others. Certain species of Medicago have 
glandular pubescence on stems, leaves, and pods ; others on the pods 
only; while some have none at all. Of the species studied the fol- 
lowing have glandular pubescence strongly developed on the pods 
at least: M. soleirolii, M. rigidula, M. minima, M. disciformis, M. 
blancheana, M. tunetana, M. faicata viscosa, and M. gaetula. Other 
species having glandular pubescence less well developed or showing 
only under certain conditions are M. murex sorentinii M. orbicularis , 
M. lupulina, and M. sativa. 

In some species of Medicago it has been noted that very minute 
glandular hairs may be present on the young green pods and dis- 
appear with later development. In other species in which only 
minute glandular hairs occur early in the season, both young and 
mature pods have well-developed glandular hairs later during hot, 
dry, and adverse weather conditions. Thus in plants in which glandu- 
lar hairs can be noted on the pods early in the season only with a 
low-power compound microscope, glandular pubescence is quite 
strongly developed later in hot, dry weather. 

Inasmuch as subspecific distinctions have been made in various 
species of Medicago on the presence or absence of glandular pubes- 
cence, the variability of this character is of value in determining the 
validity of such classification. In these studies M. lupulina and M. 
orbicularis have been especially noted. Individual plants have been 
observed throughout the season with regard to glandular pubescence 
development. 

On April 27, 191 5, at Chico, Cal., a plant of M. orbicularis was 
observed as having both young and well-developed pods that were 
not glandularly pubescent as the term is usually used but had glandular 
pubescence which could be seen with a low-power compound micro- 

1 Contribution from the Office of Forage-Crop Investigations, Bureau of 
Plant Industry, U. S. Department of Agriculture, Washington, D. C. Received 
for publication January 24, 1918. 



l60 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



scope. On May 27, 191 5, this same plant was noted as having both 
voung and mature pods with well-developed glandular pubescence, 
that is, the glandular pubescence was sufficiently well developed to be 
seen without magnification. On April 27, 1916, at Chico, Cal., the 
pods of a plant of M. orbicularis were observed as not glandular. 
The most advanced pods were well developed but not mature. A 
hand lens only was used in making this observation. On May 27, 
191 6, at which time some of the pods were mature, this same plant 
was noted as having pods not glandular, while on July 7 it was noted 
as having both young and mature pods that were very glandular. 
General observation of numerous plants of M. orbicularis and its 
subspecies in the field have shown that early in the season none of 
these has pods in any stage of development that show glandular 
pubescence, while later in the season or after hot, dry weather has 
prevailed for some time practically all plants have pods in all stages 
of development with strongly developed glandular pubescence. In 
M. orbicularis glandular pubescence has been observed in no case as 
occurring other than on the pods. 

Observations made regarding glandular pubescence in M. lupulina 
show the same fluctuating variation due to environmental conditions 
as that noted in M. orbicularis. However, there are also some indi- 
cations that in this species some forms have glandular pubescence 
throughout their existence regardless of environmental conditions 
while others are never glandular. In the common form of this 
species, wherever observed, glandular pubescence is not present early 
in the season regardless of the stage of development of the plant, 
while Liter in the season it is quite conspicuous. In M. lupulina glan- 
dular pubescence may occur on the stems, leaves, and pods, but when 
not well developed it is most conspicuous on the pods. Plants of M. 
lupulina collected at Chico, Cal., on May 18, 1909 show young pods 
glandular. Hants collected at Chico June 5, 1912 show no glandular 
pubescence. Plants collected at Chico May 6, 1914, while in general 
not glandular, show a very few glandular hairs forming. Plants col- 

ti I at New London, Ohio, June 28, [915 show pods mostly 11011- 
glandular but with few glands forming. On April 27, 191 5, at Chico, 
Cal., a plant of .1/. lupulina having pods well developed was observed 
as nol glandular. On July 15 tins same plant was noted as very 
glandular. On April 26, [916 a planl of 71/. lujmlina was observed at 
I WCO, ( al., .'J not glandular. On May <X, 1916 this same planl was 
DOted a- not glandular, while on July 13, 1916 it was noted as very 
glandular. A plant Of M. lupulina ^rowin^ at Klyria, Ohio, was ob- 



m'kee: glandular pubescence in medicago. 



161 



served on June 19, 1917 as not glandular. This same plant on August 
23 was noted as not glandular, while on October 18 it was noted as 
very glandular. 

In the summer of 1916 a plant of M. lupulina growing at Albany, 
Ore., was observed as not having glandular pubescence, while at the 
same time plants growing nearby were quite glandular. Seed from 
the non-glandular plant was collected and grown at Chico, Cal., in 
191 7. At no time did the plants grown from this seed show glandular 
pubescence. 

From observations made in the cases of M. lupulina and M. orbicu- 
laris it is very evident that the occurrence of prominent glandular 
pubescence may be due to environmental conditions, these conditions 
apparently being hot, dry weather and unfavorable soil-moisture 
supply. 

Some interesting observations have been made with regard to 
glandular pubescence in various varieties and hybrid forms of M. 
sativa and M. falcata. A large number of introductions of these 
species and subsequent hybrids made from these introductions have 
been observed to determine to what extent glandular pubescence 
might occur. Aside from one introduction of M. falcata, which had 
been identified as M. falcata viscosa, none have strongly developed 
glandular pubescence. However, glandular pubescence developed to 
some degree has been noted in a number of M. falcata forms. In 
the case of M. sativa, or common alfalfa, glandular pubescence was 
found in only two instances. In the case of M. falcata (excepting 
M. falcata viscosa), and also in M, sativa, the glandular forms are 
apparently hybrids, though they show but little variation in other 
characters. 

On April 28, 191 5 the pods of a large number of species of 
Medicago were examined to determine the presence or absence of 
glandular pubescence in species in which it might occur. Both macro- 
scopic and microscopic observations were made. In such species as 
M. scut ell ata and M. rugosa, in which glandular pubescence is always 
present, long, well-developed glandular hairs were very conspicuously 
present to the unaided eye. In other species, such as M. orbicularis, 
which has strongly developed glandular pubescence in midsummer, 
glandular pubescence was present but only microscopically. Other 
species, such as M. hispida and its subspecies, which have not been 
noted as glandular late in the season, had microscopic glandular 
pubescence on the young pods. In a number of species the glandular 
hairs per square millimeter were counted and the length of the hairs 
and the length and width of the glands were determined. 



1 62 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

In the case of M. orbicularis the maximum length of the hairs was 
63 microns, while the glands of maximum size were 36 X 20 microns. 
Later in the season the glandular pubescence of this species measures 
practically the same as M. sciitcllata, the hairs being about 600 microns 
in length and the glands about 50 microns in length. In the case of 
a number of species such as M. hispida, M. echinus, M. obscitra helix, 
M. intertexta, and M. tuberculata aculcata, microscopic glandular 
pubescence was noted on the pods early in the season, but in no case 
have these been noted as developing further. 

The number of glandular hairs per square millimeter varied con- 
siderably in the different species. In the case of M. sciitcllata and M. 
rugosa, in which species the glandular hairs were well developed, the 
number was so many that they could not be definitely counted. With 
the species having only microscopic glandular hairs the number varied 
from 40 to 120 per square millimeter. 

The data here presented seem to show clearly that in certain species 
of Medicago glandular pubescence varies decidedly with environ- 
mental conditions, and for this reason it can not be depended upon 
as a constant character in determining subspecies. 



CUTTHROAT GRASS (PANICUM COMBSII)} 
C. V. Piper. 

Panicum combsii Scribner and Ball was described 1 " from speci- 
mens collected " in damp, fertile flat woods " at Chipley, Washington 
Co., Fla., in the northern part of the State. Hitchcock and Chase 3 
record it from three localities in southern Georgia, four stations in 
northern Florida, and one station each in southern Alabama, southern 
Mississippi, and southern Louisiana, and give as its usual habitat 
"margins of ponds and wet woods." From these data it would ap- 
pear to be ;i rare species. During the past season the writer found it 
enormously abundant in Polk County, central Florida, where it is 
known as "cutthroat grass/ 1 and the peculiar areas in which it occurs 

■Contribution from the Office of Forage-( rop Investigations, lUircau of 
I'lant Industry, I' S I hpartnu nl of Agriculture, W ashington, 1). G Received 

for publication February 4. roi& 

' ' ' I l.amson . and Kill, ( ark-ton l\. Studies on American grasses. 
I' S hept. \gr., Div A«rost Mul. 24: 42-43, fig. 16. 190I. 

Hitchcock. V S.. and Chase, Agnes, The North American species of Pani- 
cum. Contr. U. S. Natl. Herb. 15: 106-107. 1910. 



piper: cutthroat grass. 



163 



in nearly pure growths are known as " cutthroats." The grass is said 
to be abundant in similar areas in Osceola County to the eastward, 
and in De Soto and Lee counties to the south. A " cutthroat " usually 
if not always occurs on seepage areas on the sides of slopes, especially 
of sand ridges. At these seepage places an abundance of water 
exudes, so that even in extremely droughty seasons, as in the spring 
and early summer of 191 7, water can be obtained at shallow depths, a 
fact utilized by stockmen. The soil of a " cutthroat " consists of a 
very fine, slippery, black muck a foot or more in depth. These areas 
are treacherous, and in spite of the heavy cover of grass wagons and 
automobiles easily become bogged. Such areas vary in size from one 
to many acres, and often successions of them occur up the side of a 
gentle slope. 

Cutthroat grass grows in dense tufts, the tough wiry leaves being 
nearly erect and 6 to 18 inches long. In typical cutthroat soil the 
plant rarely blooms. In the neighborhood of Florinda, Fla., how- 
ever, abundant specimens were found in late bloom November 4, 191 7. 

According to Mr. W. F. Ward, the superintendent of the Kissim- 
mee Island Cattle Company, the stockmen of the region are in accord 
as to their experience with this grass as a forage. They consider it 
good fattening winter feed for adult steers and for non-pregnant 
cows, but that pregnant cows abort when pastured on this grass, and 
that young animals die. So fixed is this opinion that stockmen govern 
their operations accordingly. Locally the disease is known as " salt 
sickness." " Salt sickness " has several times been investigated in 
Florida. Stockbridge, French, and Ennis 4 summarize their investiga- 
tions as follows : 

1. The disease known locally as " salt sickness " is not believed to be a specific 
disease, but rather a condition resulting from improper environment, especially 
insufficient nutrition. 

2. Similar occurrences have existed elsewhere and are usually confined to 
regions where the predominating soil is light, sandy, comparatively lacking in 
nutritious qualities, Cape Cod peninsula in Massachusetts being a locality 
similarly affected. 

3. The condition is most prevalent at the end of the winter season, when 
animals have been for several months confined upon range or pasture consist- 
ing of the dry wire-grass and other inferior vegetation of the sand ridge por- 
tion of the State. 

4. The disease is distinctly digestive in character, has its seat in the alimen- 
tary canal and finally develops into chronic inflammation of the small intestine, 
resulting in malnutrition, anaemia and frequent death from starvation. 

4 Stockbridge, H. E., French, W. E., and Ennis, J. E. Salt sickness. In 
Fla. Agr. Expt. Sta. Rept. 1900-01, p. 43-58. 1902. 



164 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



5. The symptoms show those generally attendant upon chronic anaemia : 
loss of flesh, loss of appetite, or abnormal appetite, including craving for 
foreign substances like earth, sand and bone, and diminishment of red blood 
corpuscles, as evidenced by thinness of blood, swelling or ulceration under the 
jaw and white bloodless appearance of mucus membranes particularly inside 
of the mouth and eyelids. 

6. Alimentary correctives and tonics are suggested as counteracting these 
conditions. 

7. The change of affected animals to new range or pasture is both pre- 
ventive and curative in effect. 

8. The use of lime water and gentian or iron salts have proved invariably 
beneficial and during our investigation of the subject not an animal thus 
treated died, but all eventually recovered normal condition. Air-slacked lime 
2 ounce, and sulphate of iron h ounce, are recommended in 3 gallons of water, 
the former as often as the animal will drink and the latter daily. 

Similar conclusions are reached in a separate publication by Dr. W. 
E. French. 5 Dr. C. F. Dawson, 6 however, reaches a different con- 
clusion : 

From what I have seen of "salt-sick" and from what we know of Texas 
fever, the role which it should play in diseases of cattle in the South, and 
especially in Florida, where the tick is ever-present, I am forced to believe 
that " salt-sick " is chronic Texas fever and that the conditions named by Dr. 
Stockbridge as being the cause of the disease are not the sole causes, but are 
contributing causes. The loss of appetite which occurs early in "salt sick" 
cannot in all cases be attributed to poor pastures and it would occur, regard- 
less of pasture conditions, in an animal attacked with Texas fever. 

In all cases of " salt-sick " where I have had an opportunity of making a 
post-mortem examination, I have found the appearances to be those which are 
attendant upon the extreme anaemia which follows an attack of Texas fever. 
Tl ' ■ are pale, watery blood, dropsical conditions, light-colored, bloodless 
liver and extreme emaciation. Most important of all was the occurrence in 
fairly large numbers of the germ of Texas fever in the red blood corpuscles, 
in the omental fringes, and of ulceration of the pyloric end of the fourth or 
true stomach. 

I he subject BO far as "cutthroats" are concerned is one that merits 
can ful investigations as the alleged facts seem to be connected with 
L^rass. Botanically the subject is of interest because a grass 
otherwise rare i- enormously abundant in the peculiar soil areas to 
which it is adapted. 

•French, W. K. A study of salt-sick cattle. /;/ Amer. Vet. Rev., 25: 985- 
901. igoz 

■Dawson, C F. Texai Cattle fever and salt sickness. Fla. Agr. Expt. Sta. 
I Jul. 64. 1902. 



hill: a drill for nursery seeding. 165 



A DRILL FOR SEEDING NURSERY ROWS. 1 

C. E. Hill. 

In view of the large amount of nursery sowing of field crops done 
on experiment stations, any implement which will save time and labor 
and eliminate mixtures of varieties or strains in seeding is of value. 
A drill which has these advantages has been devised by the writer. 
The merits claimed for the drill include the following : 

1. Greater speed. 

2. Greater accuracy. 

3. Operation by one man. 

4. Elimination of mixtures. 

5. Seeding can be done on a windy day. 

6. Better germination can be obtained than in hand seeding. 

The essential parts of the drill are a funnel into which the seed is 
dropped by hand at the desired rate : a furrow opener ; a tube which 
carries the seed from the funnel to the furrow opener ; and a carriage 
on which these parts are mounted. The tube is long enough to permit 
the operator to walk with the body erect while seeding and wide 
enough to prevent any seed from lodging. 

The type of carriage can vary somewhat and yet be satisfactory. 
A carriage having three wheels, two in front and one in the rear 
behind the furrow opener, is recommended. This type of carriage 
will stand without support and can be guided easily. 

The front wheels are adjustable on the axle so that when the dis- 
tance between the rows to be planted is not greater than 1^2 feet 
they can be made to mark the rows by adjusting the distance between 
the wheels. After sowing the first row the other rows will be spaced 
properly by running one front wheel on the row previously sown. 
Rows more than i 1 /* feet apart can be marked by having an extension 
rod marker attached to the drill. 

The drill is pushed by the body in contact with a padded curved 
band, attached to the frame of the drill at about the height of one's 
waist line. It is guided by one hand on the frame in front of the 
body. The other hand is used in dropping the seed into the funnel. 

1 Contribution from the Office of Forage-Crop Investigations. Bureau of 
Plant Industry. U. S. Department of Agriculture. Washington, D. C. Received 
for publication January 24, 1918. 



I 66 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

The depth of seeding is regulated by an extension sleeve on the 
furrow opener. 

Seed in envelopes for sowing 100 or more varieties can be carried 
in a box conveniently placed on the frame of the drill. If a seeding 
outline is made and the envelopes are arranged in order of sowing, it 
is not necessary to label the rows until the seeding is completed. 

The drill is recommended for sowing short nursery rows of dif- 
ferent varieties or strains. With a little practice the rows can be 
seeded as fast as the drill can be pushed over the ground, as no time 
is lost in calibrating or cleaning the drill. If the envelope containing 
the seed is opened in the funnel the possibility of mixtures is elimi- 
nated, as the seed is carried directly to the ground in the row to be 
seeded and none can lodge in the drill. As drilling disturbs the sur- 
face soil but little, the seed usually can be drilled into moist ground 
at a uniform depth, insuring better germination than can be obtained 
by hand seeding into an open furrow. The drill is suitable for sowing 
light and fluffy seeds like tall oat grass {Arrhcnatherum clatius), 
which will not feed through any of the ordinary types of garden drill. 
The drill is very simple in construction and can be made at a very 
low cost. 

From these plans a drill essentially the same as the one described 
was made by Mr. F. J. Schneiderhan at the experiment station, Moro, 
Oregon, in 191 7, from material available on the station. This was 
found satisfactory for sowing all kinds of seed and was used for all 
short-row seedings except where definite spacing of the seed was de- 
sired. It was also used in sowing long rows of light, fluffy seed that 
could not be seeded with a Planet Jr. garden drill. In his wheat 
nursery seeding, Mr. Schneiderhan was able to sow 350 rows 5 feet 
long per hour. Only 350 rows could be sown in a day by the method 
formerly used, which was to mark the rows to be sown, open the 
furrow, drop the seed by hand, and fill the furrows, making in all 
four Operations. At the Moro station this drill was especially liked 
on account of its elimination of possible mixtures. 



SPRAGG I RED ROCK WHEAT AND ROSEN RYE. 



167 



RED ROCK WHEAT AND ROSEN RYE. 1 

Frank A. Spragg. 

I wish to take a little of your time this afternoon in discussing some 
of the new products of the plant breeding work at the Michigan Agri- 
cultural Experiment Station. 

When I took up the work ten years ago, our highest yielding wheats 
were white wheats of poor milling and of especially poor baking 
quality. The problem was to find a red wheat that would at least 
equal the white wheats in yield. This we are finding in the Red 
Rock, which originated from an individual plant selected in 1909. 
This strain was grown in a row in 1910, planted in the regular varietal 
series (twentieth-acre plat) in the fall of 191 1, and distributed in 
peck lots through county agents in the fall of 191 3. The Red Rock 
is a red wheat of exceptional winter hardiness, high yielding ability, 
an extra stiff straw, and those characteristics which make a bread of 
unusual quality. 

The Rosen rye, on account of the exceptionally poor competitors, 
is yielding about twice as much as common rye. It has a shorter, 
stifTer straw than common rye and much larger heads, which are ex- 
ceptionally well filled. There are four rows of kernels on every rye 
head, but the common rye has only scattering kernels along each row. 
The Rosen has four very nearly complete rows on 99 percent of its 
heads. 

While I am talking about new pedigreed grains I want to mention 
the Michigan Winter barley. It has not proved as popular as the 
others, because it apparently does not fit into the rotation as well. It 
must be sown between August 15 and September 10 in order to pro- 
duce a root system sufficient to stand the winter and give a good yield 
the next summer. When planted early on well-prepared, fertile soil, 
yields as high as 64 bushels to the acre have been reported. It ma- 
tures before the wheat at a season of the year when the farmers 
usually need the grain to fatten their hogs. Unfortunately, however, 
on September 1 almost no land is available for the sowing of a fall 
crop. 

1 Contribution from the Michigan Agricultural Experiment Station, East 
Lansing, Mich. Presented by the writer, with illustrations, at the tenth annual 
meeting of the American Society of Agronomy, Washington, D. C, November 
13, I9I7- 



1 68 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



You may have heard of these pedigreed grains to some extent. 
They are spreading rapidly. Inquiries come to us from nearly 
every State in the central and eastern portions of the country, ex- 
tending as far west as New Mexico, where a county agent has pur- 
chased the Red Rock wheat for sowing by his farmers this fall. 

The distribution of the Rosen rye was started in Michigan in 1912, 
when about six bushels were distributed, mostly in bushel lots. It 
seems that we have to distribute to a number of people before we find 
one that will make the best use of a new thing, and in this case it 
seems that Carlton Horton, of Albion, was the only man who really 
took advantage of his opportunity. At that time pedigreed grains in 
Michigan were new and very little known. Mr. Horton planted his 
bushel on an acre and the next year got 35 bushels. The man who 
thrashed the rye and the neighbors who helped became very enthusias- 
tic, so Mr. Horton thought he might easily have sold the rye at $5 a 
bushel for seed if he had cared to do so. This rye has spread through 
his own county, Jackson, until now the county agent tells us that 
almost no common rye is to be found. It has also very nearly re- 
placed the common rye in St. Joseph County. It is estimated that 
15,000 acres of Rosen rye were grown in Michigan in 191 7 and 230,- 
000 acres sown for the season of 1918. 

The distribution of Red Rock wheat began in 191 3 by the sending 
out of peck samples to a number of county agents. I will just men- 
tion a few samples and results. The peck that went to Allegan 
County was sowed on the farm of John Odell, about six miles south 
of Allegan, on his garden patch in 1914. It produced at the rate of 
55 bushels to the acre. He sowed six acres that fall and in 191 5 had 
seed for sale. Though he and County Agent Cook advertised it, the 
people knew it was a bearded wheat and did not buy until some of 
the seed had been ground by the local miller, who then became en- 
thusiastic. The favorite trick of the miller was to have two bushels, 
wrl] cleaned and closely tied in a sack, sitting near so that when a 
fanner came in the miller invited him to lift the sack. When Mr. 
Parmer did not lift it at arm's length, he inquired what it was. The 
miller told him thai il was the only wheat grown in Michigan from 
which he could really make good bread, and on the recommendation 

of the miller the farmei owed the seed and kept it pure until in 1917 

*ai JOO acres of tin- wheal in Allegan County still pure enough 
to pass tlx inspection as pedigreed grain. 

The peck ample that went lf» Kent County became mixed, and 
though there wen 100 ;■< re- mown in 1917 it could not be used for 



spragg: red rock wheat and rosen rye. 169 

pedigreed grain. The peck sample that went to Newaygo County in 
1913 has been cared for by the county agent and in 1916 had been 
increased to 700 bushels. Four hundred bushels of this went to 
Kent County to replenish their supply of pure seed. I will just men- 
tion one other county. That is Houghton, on the Upper Peninsula, 
reaching into Lake Superior. Mr. Geismar is county agent there. 
During the years 1914 and 191 5 he acclimated this wheat so that in 
1916 it produced a fine crop on the county farm at Houghton. From 
this seed fields were grown in Houghton, Ontonagon, Marquette, and 
Delta counties in 191 7. Thus, the Red Rock wheat when sown 
August 15 is proving successful in a spring wheat district. 

It is estimated that 4,000 acres of Red Rock wheat were raised in 
the State in 1917 and that 100,000 acres were sown for the 1918 
crop. The demand for this grain has been so great that just at the 
end of the campaign I received a letter from a farmer who stated 
that he had failed to obtain any Red Rock for seed and wished to 
have his name put on the list to receive some next season. 

A representative of the Federal Bureau of Markets, Mr. Frank, 
who was looking up samples of our Michigan wheats in the fall of 
1917, saw a large number of samples of Red Rock that had been 
sent in by the various growers as samples that they guarantee to be 
just what they were selling. Some of these samples weighed as 
high as 64 pounds to the bushel, while others had not been allowed 
to pass the grain inspection because of light weight, smut, or mix- 
tures of other varieties. Mr. Frank obtained nine 2-bushel lots 
from farmers, ranging from the best to the poorest, for a milling test. 
He is of the opinion that a special grade will be given to Michigan 
into which only Red Rock is good enough to fall. 

The distribution of our pedigreed grains has been aided greatly 
by the county agricultural agents, working in connection with the 
Michigan Crop Improvement Association, of which Mr. J. W. 
Nicolson is secretary. The members of this association are in gen- 
eral the most up-to-date and progressive farmers in their districts. 
Selected members of the association receive small samples of new 
crop varieties from the experiment station. There are also general 
demonstrations that are offered to members, for which the associa- 
tion must buy the seeds. The members get the benefit of the ad- 
vertising of the association and the sale of pedigreed seeds when 
they have any that are acceptable. The expenses of the association 
are, first, the cost of the seeds for demonstration purposes, except 
such as are furnished by the plant-breeding work. Because of the 



I/O JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



size of the association, this latter is simply a start. Other expenses 
include the salary and the traveling expenses of the inspector, the 
advertising in State, county, and agricultural papers, and the print- 
ing of seed lists of growers. There is no State aid for this distribu- 
tion work as there is in Wisconsin. All the expenses must be paid 
out of the membership and inspection fees, except for the printing 
of circulars informing the farmers of the benefits derived from the 
raising of these grains. 

The membership fee is $i a year. If the individual is* isolated the 
inspection of pedigreed grain costs him $10, of which he sends $6 
with his application for inspection and pays the remainder when the 
field is inspected, if it is allowed to pass. If three or more members 
can be reached in a single day by the inspector, then the inspection 
fees are reduced by half. The farmer who is really benefitted by 
the raising of pedigreed grains must sow the new grains on clean 
land, with a clean drill, out of clean sacks. He must treat the grain 
for smut. He must pull weeds such as vetch and cockle while in 
bloom ; in fact, he must remove all weeds that can not be screened 
out after thrashing. He must pull out all mixtures of other grains, 
such as rye in wheat, and if his wheat contains more than I percent 
of other varieties these must be removed, according to the require- 
ments of 1 91 7. The inspection is going to be more and more strict 
as the time goes on. The farmer must apply for inspection and en- 
tertain the inspector unless he is being taken care of by a county 
agent. 

The farmer must clean his binder and thrasher to prevent mix- 
ture-. Fall grains like wheat or rye should be thrashed after oats. 
Most of the oats will be cleaned out in fanning, while those that 
remain will be frozen out during the winter. That is one of the 
advantages Michigan has over'The West and South, where spring 
grains often volunteer. In thrashing oats after wheat, it is well to 
watch until no more wheat can be seen and then discard a few more 
bushels to be sure. If the thrashing machine is not cleaned in this 
way then at least 25 bushels must be discarded in saving seed for 
MUe. Of OOttrsei tlO one stands over the farmer with a stick to 
make him do this, hut these are the general rules that we feel must 

be obeyed or hii grain will not be able to pass the grain inspection. 

He mutt fan bis grain at home, for if he sends it to an elevator he 
iVt- it mixed beyond redemption. He must remove all weed seeds, 

all -mm hall-, and shriveled kernels thai will not produce thrifty 

'Ib'-n he mu-t M-nd the secretary of the association a peck 



spragg: red rock wheat and rosex RYE. 171 

f 

sample of the grain just as he expects to sell it. If this grain is to 
pass the second inspection it must be free from weeds, up to standard 
weight, and contain less than I percent of mixture of other varieties. 
If his grain is passed he receives the association's shipping tags and 
report cards. His name is put on printed lists of growers who have 
seed for sale. The association does the advertising and those who 
inquire as to where they can purchase good pedigreed grains are 
sent these printed lists. The growers also certify on the back of the 
shipping tags. For instance, if it is Red Rock, " The wheat in this 
bag is Red Rock ; it grew on a field inspected by an agent of the 
association; it conforms to the State seed law, and it conforms to 
sample submitted to the association for inspection." This certifica- 
tion is the purchaser's guarantee that the seed is as it is supposed -to 
be, and is binding upon the grower to the extent that he must refund 
the money or make it right. Several did this last fall. For instance, 
one grower in Allegan County refunded Si a bushel on some seed 
that had evidentlv not been recleaned. 



THE IDENTIFICATION OF VARIETIES OF OATS IN 
NEW YORK. 1 

E. G. Montgomery. 

During the summer of 191 7. Air. George Stewart, one of the 
graduate students in farm crops at the New York State College of 
Agriculture, undertook as a part of his graduate work to make a 
practical test of Dr. Etheridge's key to oat varieties 2 under field con- 
ditions. Samples of oats were collected from New York State seeds- 
men and grown in the garden for identification, and also oats grow- 
ing in fields were examined and identified. In addition, varieties 
growing at two experiment stations were also examined. Some of 
the results as worked out by Mr. Stewart are here given. 

The varieties offered by certain seedsmen and their identification 
were as follows : 

1 Contribution from the Department of Farm Crops of the College of Agri- 
culture. Cornell University. Ithaca, New York. Presented at the tenth annual 
meeting of the American Society of Agronomy. Washington, D. C, November 
13, 1917. 

2 Etheridge. W. C. A classification of the varieties of cultivated oats. X. Y. 
(Cornell Univ.) Agr. Expt. Sta. Memoir 10, 1916. 



l 7 2 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Seedsman's name. Identification according to key. 

Seedsman A 





. . .Winter Turf. 




. . . Swedish Select. 




...Do. 




...Do. 


Long's White Tartar 


. . .Tartar King. 


Imported Black Tartarian . . 


. . . Black Tartarian. 




. . .Swedish Select. 




...Do. 




. . . Storm King. 




. . . Storm King or Tartar King. 


Seedsman B 






. . . Tartar King. 




. . . Swedish Select and Storm King. 


Seedsman C 






...Lincoln and Belyak. 


Seedsman D 






. . . Swedish Select. 


Seedsman E 




Alberta Cluster 


. . . Swedish Select. 


Seedsman F 






. . . Swedish Select and Storm King. 




. . . Storm King and Swedish Select. 




. . . Storm King and Swedish Select. 


Seedsman G 






. . . Swedish Select. 


Seedsman H 








Twentieth Century 


. Belyak. 


Seedsman I 






Belyak. 



Where two varieties were identified, the dominant type is named 
first. Out of 10 varieties offered by Seedsman A, 5 appeared to be 
Straight Swedish Seleet, if the identification is correct. When this 
identification was presented to the seedsman, he stated that four of 
the varieties identified as Swedish Select were imported from Eng- 
land, which would indicate that in Europe also they are using several 
Dame! Eor what appears to be the same variety. Of course, there may 
be difference! in adaptation or quality not discernible in a botanical 
ruination, but at least to all outward appearances the varieties 
arc identical. 

Of the 22 varieties listed above, 11 are of Swedish Select type and 

1 mor< contain an admixture of Swedish Select. This would indicate 

that this type has been found besl for the State. 

II.' varietje. ^rown at one experiment station, together with the 

Identification! found, were as follows: 



MONTGOMERY : OAT VARIETIES IN NEW YORK. 



173 



Experiment station name. Identification according to key. 

1. Silvermine Silvermine. 

2. White Russian Swedish Select. 

3. Corn Belt Lincoln. 

4. President Lincoln and Swedish Select. 

5. Wideawake Irish Victor. 

6. Napoleon Swedish Select and Lincoln. 

7. Siberian Swedish Select, Lincoln and June. 

8. Burt i Burt. 

9. White Probsteier Lincoln and Swedish Select'. 

10. Golden Fleece Swedish Select. 

11. White Plume Storm King. 

12. Joanette Joanette. 

13. Canadian Side Storm King, Lincoln and Belyak. 

14. Hvitling Lincoln and Belyak. 

15. Lincoln Lincoln and Silvermine. 

16. Sixty Day Kherson. 

17. Welcome Silvermine or Scottish Chief. 

18. Czar of Russia Swedish Select, Lincoln and Silvermine. 

19. Black Mogul Black Norway (?) 

20. Clydesdale Silvermine and June. 

21. Big Four Belyak. Silvermine, and June. 

22. Seizure Green Mountain. 

23. Alaska Lincoln, Belyak, and June. 

24. American Banner Belyak and Lincoln. 

25. Morganfeller Silvermine, Lincoln, and Belyak. 

26. Swedish Select Swedish Select. 

27. Long's White Tartar Storm King. 

28. Golden Rain Golden Drop. 

29. Sparrowbill White Tartar and Storm King. 

30. Improved American Lincoln and Swedish Select. 

31. Beardless Probsteier Awnless Probsteier. 

32. Green Mountain Lincoln. 

33. Ellwood Belyak and Silvermine or Scottish Chief. 

34. Victory Lincoln and Silvermine. 

35. Early Champion Irish Victor. 

36. Sensation Lincoln, Swedish Select, and Early Gothland. 

37. Twentieth Century Lincoln, Swedish Select, and C. I. 602. 

38. Storm King Storm King. 

39. White Ligowa Swedish Select. 

40. Detmer's New Bumper Storm King. 

The above list is given to illustrate the situation that probably 
exists at all of the experiment stations where varietal tests are under 
way and emphasizes what has heretofore been pointed out, the need 
of a careful study of varietal nomenclature and the standardization 
of names. 

In order to get some idea in regard to the type of oats grown by 



174 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

New York State farmers, examination was made of a strip of 
country across the State from north to south, which is believed to be 
fairly representative. In all, 418 fields were examined with the re- 
sults shown- in Table I. 



Table i. — Number of farms on which certain varieties of oats are grown in 
Cortland, Tioga, Cayuga, Tompkins and Ontario counties, New York. 



\ ariety. 


Virgil, 
Cort- 
land Co. 


Owego. 
T. ioga 
Co. 


Auburn, 
Cayuga 
Co. 


Mo- 
ravia, 
Cayuga 
Co. 


Ithaca 
to Dry- 
den, 
Tomp- 
kins Co. 


Ithaca to 
Trumans- 
burg, 
Tomp- 
kins Co. 


Western 
Ontario 
Co. 


Eastern 
Ontario 
Co. 


Total 


Lincoln 


7 


IS 


15 


17 


7 


31 


18 


8 


Il8 


Swedish Select . . . 


11 


6 


10 


20 


9 


15 


30 


9 


110 


Belvak 


3 


3 


5 


. I 


3 


20 


9 


3 


47 


Silvermine 


7 


8 


3 


I 


8 


5 


17 


8 


.57 


Canadian 


4 




14 


7 


3 








28 


Storm King 


4 


10 


4 


1 


2 


3 


8 


2 


34 


Sparrowbill ...... 






1 


3 










4 


June 




1 




I 








1 


3 


Old Island Black . 








I 










1 


Irish Victor 






2 




1 




1 


3 


7 


Tobolsk 






1 




I 








2 


Danish Island . . 






3 












3 


White Tartar . . 












2 




2 


4 


Total 


36 


43 


58 


52 


34 


76 


83 


34 


418 



The figures given in Table 1 show the following percentages of 
the different oat varieties in central New York State : 



Variety. Per Cent. 

Lincoln 28.2 

Swedish Select 26.3 

Silvermine 13.6 

Belyak 11.3 

Storm King 8.1 

Canadian 6.7 

All others 5.8 

100.0 



The summaries show that 79.4 percent of the oats were identified 
a- Lincoln. Swcdi-h Select, Hclyak, and Silvermine. These varieties 
are very similar and can hardly be distinguished from each other 

except bj a trained observer. The indications are that long experi- 

eno of both the Seedsmen and farmers have shown this type to lie the 
1m--i adapted to the Stan 



coe: early varieties of velvet beans. 



175 



ORIGIN OF THE GEORGIA AND ALABAMA VARIETIES OF 
VELVET BEAN. 1 

H. S. Coe. 

For many years the Florida velvet bean {Stizolobium deeringianum 
Bort) has been grown in Florida and the extreme southern part of 
the Gulf States as a soil-improving crop and for grazing. In other 
portions of the South the value of this plant was limited, as it re- 
quired a frost-free season of eight to nine months to nature. Even 
though only a portion of the pods usually matured in the southern 
half of the Gulf States, many farmers valued this crop so highly for 
grazing and for soil improvement that they planted it annually. 

As the United States Department of Agriculture recognized the 
value of an early-maturing velvet bean which would produce winter 
grazing equal to that of the Florida variety and which would mature 
in most parts of the cotton belt, a careful search was made for such 
a plant in other countries. The Chinese velvet bean was introduced 
from Tehwa, China, and the Yokohama velvet bean from Yokohama, 
Japan. Both of these plants mature earlier than the Florida variety 
but their pods have the undesirable characters of splitting and shatter- 
ing the seed when mature. The Chinese variety is superior to the 
Florida for the southern portion of the Gulf States, but it rarely 
matures in the northern part of the cotton belt. The Yokohama 
matures in about 120 days, but it produces a small vine growth and 
most of the pods are formed so close to the ground that in a rela- 
tively short time they decay. However, before either of these species 
was introduced, two early maturing mutants of Stizolobium deer- 
ingianum were found in southern Georgia, although they were un- 
known to the public for several years. 

In the spring of 1906, Mr. Clyde Chapman of Sumner, Ga., planted 
a field of corn and Florida velvet beans. The following August, 
several hills of mature beans were found in this field. The seed 
collected from these early-maturing plants was planted in corn the 
following year and early plants were produced, similar to those found 
the previous year. A small quantity of the seed grown in 1907 was 
distributed to several of Mr. Chapman's neighbors, but so far as 

1 Contribution from the Office of Forage-Crop Investigations, Bureau of 
Plant Industry, U. S. Department of Agriculture, Washington, D. C. Received 
for publication January 24, 1918. 



1/6 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



known none of this seed was sent out of this immediate .section prior 
to 1 91 2, with the exception of one lot sent to Schley County, Ga., 
in the spring of 1909. The seed of the Florida velvet bean planted 
in the spring of 1906 was obtained from Mr. Z. C. Allison, of Sumner, 
Ga., a relative of Mr. Chapman's. Three or four years prior to this 
date Mr. Allison obtained seed of this particular strain from a field 
in Schley County, Ga., and each year enough seed matured on his 
farm for seeding the following spring. 

The first early-maturing plants noted by Mr. Chapman made a 
small vine growth, matured their first pods in 90 to 100 days from 
planting, and produced seed which, according to Mr. Chapman, was 
somewhat paler and slightly smaller than the .seed of the Florida 
velvet bean. From the first the plants of this early-maturing variety 
came true to type. 

Another mutant of the Florida velvet bean was found by Mr. R. 
W. Miller, of Broxton, Ga., in August, 1908, in a field planted to 
corn and Florida velvet beans. The velvet bean seed planted in this 
field was grown in Florida. The seed of these early-maturing plants 
was sown in 1909 and, according to Mr. J. D. Harrell, of Douglas, 
Ga., a relative of Mr. Miller's, the plants came true to type. Mr. 
Harrell obtained some seed from Mr. Miller in the spring of 1910 
and that year harvested 15 bushels. So far as known, the seed of 
this mutant was not generally distributed by Mr. Miller or Mr. 
Harrell, but Mr. W. A. Clark, of Jacksonville, Ga., obtained seed 
from Mr. Harrell in 191 3, and later sold seed of this variety under 
the name of " Clark's velvet bean." This bean, so far as can be 
-mined, is identical with the one found by Mr. Chapman. In 
tigating the origin of this mutant, not the slightest evidence was 
found to indicate that the early-maturing plants found by Mr. Miller 
could have been produced from seed grown by Mr. Chapman. 

\ third mutant of the Florida velvet bean was found by Mr. H. 
L Blount, of Flomaton, Ala. Mr. Blount, who had grown velvet 
a grazing crop for at least fifteen years, obtained seed 
from Florida, which he planted with corn in the spring of 1 91 1 . In 
a sandy gap in this field, one plant was observed which bloomed 
much earlier than any Oi the others. By October 20, and before 
fro 1 all of the seed on this plant was matured and the plant was 
dead. tbOttl 1 pintl Of Seed were collected from it. This seed 
Wrai planted along a fence row in [912, and that fall several bushels 
of seed were harvested. In the spring of 1913, 50 to 60 packages 
of thil iced, varying in rize from one half pint to one quart, were 



coe: early varieties of velvet beans. 



l 77 



sent to farmers in Alabama, Mississippi, and Georgia for trial, and 
the remainder of the 191 2 crop was planted in corn on Mr. Blount's 
farm. In 1914 seed of this variety was sold to a number of Ala- 
bama farmers and to the Alabama Agricultural Experiment Station. 
Mr. Blount stated that as soon as he found the early-maturing plant 
he wrote to the person in Florida from whom he purchased his 
seed, asking him if he had observed early-maturing plants in his 
fields. The reply stated that an early-maturing velvet bean had 




Fig. 25. Map showing the extension of the velvet bean area by the introduc- 
tion and discovery of early maturing varieties. The Florida velvet bean will 
seldom mature fully north of line No. 1, while the Georgia variety will mature 
south of line Xo. 2. 

never been observed by the writer. The early-maturing plant found 
in 191 1 made a somewhat smaller vine growth and produced slightly 
smaller and somewhat paler seed than the Florida velvet bean, but it 
made a larger vine growth than the early-maturing varieties found 
by Mr. Chapmen and Mr. Miller. This variety matured 1 fully in 
170 to 180 days, or about two months earlier than the Florida velvet 
bean. It is known as the Alabama velvet bean. 

The N. L. Willet Seed Company, of Augusta, Ga., the Office of 
Forage-Crop Investigations, the Office of Extension Work in the 



I78 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



South, and Mr. Clifford Grubbs, manager of the Farmers' Produce 
Exchange, of Sylvester, Ga., were largely responsible for the gen- 
eral distribution of the early variety discovered by Mr. Chapman. 
Mr. YYillet purchased about 200 1 bushels of seed of this variety from 
a neighbor of Mr. Chapman's in the fall of 191 4. This seed was 
sold to the seed trade in the spring of 191 5 under the name of 
" Hundred-Day Speckled." Twenty bushels was purchased from the 
N. L. Willet Seed Company by the Office of Forage-Crop Investiga- 
tions in the spring of 191 5 and distributed in small packages in 
different sections of all of the Southern States in 191 5 and 1916. 
The Office of Extension Work in the South purchased a quantity 
of seed of this variety at Atmore, Ala., in the spring' of 1915. Some 
of this seed was sent to each county agent in Alabama and Missis- 
sippi, and to a few agents in Louisiana, eastern Texas, and South 
Carolina for planting on demonstration farms. Mr. Clifford Grubbs 
shipped to many points in the South the surplus grown by the 
farmers living in the vicinity of Sumner. The first shipments were 
made in 1913 and about 25 bushels were sold. In 1914 approxi- 
mately 100 bushels and in 191 5, 500 bushels were distributed in 
this manner. 

The Office of Forage-Crop Investigations has suggested that the 
name " Georgia " be used in preference to " Hundred-Day Speckled " 
because the variety was first discovered in Georgia and further 
because it does not mature in 100 days, as the name " Hundred-Day 
Speckled " indicates. The first few pods may mature in that time 
but it usually requires 120 to 130 days for the entire crop to ripen. 

It is probable that early maturing mutants of Stizolobium deerin- 
gianum have appeared elsewhere, but, if observed and isolated, no 
records have been obtained. 

As both the Georgia and Alabama varieties have been distributed 
to nil parts of the South, it is impossible to say in what percentage 
each contributed to the 1917 crop. As the Georgia variety was more 
generally distributed in 191 5 and as much of the seed grown that 
year was sold for Beeding purposes, it is assumed that the acreage of 
Georgia was much larger than that of the Alabama in 1916 and 
\<)ij. It is believed that the Alabama variety will be planted most 
exteti ively in the southern portion of the cotton belt and the Georgia 
variety in the northern portion on account of the time required for 

each to mature. The Alabama variety makes a larger growth and 
fore hould yield more heavily than the Georgia in sections 
ere the growing season is long enough for it to mature. 



coe: early varieties of velvet beans. 



179 



By the discovery and distribution of these early-maturing varieties 
the culture of velvet beans has been greatly popularized and has ex- 
tended to the northern limits of the cotton belt (fig. 25). Not only 
has the area of adaptation been largely increased but the acreage of 
velvet beans has increased from less than 1,000,000 to over 5,000,000 
in the past three years (fig. 26). The acreage in 191 7 was 119 
percent greater than in 1916. The direct and indirect value of this 
enormously increased acreage of a vigorous-growing legume will be a 
determining factor in improving the agricultural industry of the 
South, as the large quantity of nutritious feed produced by this crop 
at a low cost will stimulate the production of livestock. The net cash 




Fig. 26. Map showing the distribution of velvet beans. 



value of the early-maturing velvet beans produced in corn was more 
than $20,000,000 in 1 91 7. To the net cash value of the beans and 
pods may be added the value of the increased yields of subsequent 
crops, as the vines will be plowed under or pastured and when pastured 
but little of their fertilizing value is removed from the field. Experi- 
ments conducted by different experiment stations show that velvet 
beans are superior to cowpeas, beggarweeds, or soybeans for improv- 
ing the soil. Therefore, it is believed that the value of the vines and 
roots of the 191* 7 crop for this purpose is at least twice as great as 
the cash value of the beans and pods. 



I SO JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



THE VALUE OF BLUE LITMUS PAPER FROM DIFFERENT 
SOURCES AS A TEST FOR SOIL ACIDITY. 1 

P. E. Karraker. 

Many workers in soils, through publications and otherwise, advise 
that farmers use the w T ell-known blue litmus paper test as a qualita- 
tive field determination for soil acidity. A smaller number express 
doubt as to any great value attached to this test, especially in the 
hands of farmers, and some few think that the farmer should leave 
the test entirely alone. Barlow 2 has gotten together a number of 

In the soils laboratory of the agronomy department of the Ken- 
tuck}' station it is the custom to use the blue litmus paper test to get 
an initial idea as to the reaction of the soil before employing the 
quantitative lime-water determination, and to a limited extent the test 
is used independently when only qualitative information is desired. 

Kalbaum's blue litmus paper, a stock of which had been on hand 
since the fall of 191 5, was used in this work. This paper is very 
sensitive and apparently has nearly always given accurate qualita- 
tive results. In addition it has afforded some information in a quan- 
titative way. Recently, blue litmus paper tests were made in 32 soils. 
Afterwards, occasion arose to determine the limestone requirement of 
these soils by the Hopkins and also by the Veitch lime-water methods. 
In but one or possibly two instances was the qualitative information 
afforded by the litmus paper test at fault. 

Lately in connection with the preparation of a practical bulletin on 
liming land, the question arose as to the proper time to leave the blue 
litmus paper strips in the soil. Reference to the literature showed a 
considerable diversity of direction on this point. For instance, Whit- 
Bon and Weir 1 advise that the paper be left in 5 minutes; Abbott 4 
about 10 minutes; Sehollcnberger 5 about half an hour; and Barker 
and I'aer'' as much as half an hour. 

1 Contribution from the Agricultural Experiment Station, University of Ken- 
tucky. Lexington, Ky. Received for publication December 13, 1917. 
In Jour. Amor. Soc. Agron., X: jf), 27. 1916. 

■ Barlow, J. T. Soil acidity and the litmus paper method for its detection, 
tli*- published opinions. \ 

1 WhitSOU, U. A , and Weir, W. W. Soil acidity and liming. Wis. Agr. 
Kxpt Sta. Mul. 230, p. 9. 1913. 

I Abbott, John M. Liming tin- soil. Ind ( Purdue Univ.) Agr. Expt. Sta. 
Circ. 33, p. 12. 1912. 

< r. ( . J. Acid soils and sod acidity. /;/ Ohio Agr. Expt. Sta. 
Monthly Mul.. p. 33. October, 191 7. 

B rllCr, II and M;i»t, VY W. Ground limestone for use in New York 
Statr. N. Y. State Agr. Kxpt. Sta. (Geneva) Bui. 430, p. 29. 



karraker: litmus paper tests for acidity. 



181 



It has been the custom in this laboratory to leave the litmus paper 
in the saturated soil only 5 to 10 minutes, usually only 5 minutes. 
In nearly all cases a distinct pink and often the maximum pink occurs 
in this time when the soil is acid. The color change occurring in a 
much longer period, 30 minute's or more, is not considered as trust- 
worthy as that from the shorter period, as there is a tendency for a 
pink color to develop in this longer period whether the soil is acid 
or not. 

Obviously the length of time that the paper should remain in the 
soil will depend on its sensitiveness. Litmus paper from different 
sources varies much in this respect. The worker is very likely to 
recommend the length of time best adapted to the paper with which 
he is familiar. If it were not. for the tendency of a pink color to 
develop in time, particularly in heavy soils, even in the absence of an 
acid condition, a maximum time could be prescribed which would 
enable the color change to occur with litmus paper of all grades of 
sensitiveness. The fact that litmus paper from different sources does 
vary in sensitiveness and that a maximum time can not be prescribed 
on account of danger of appearance of a pink color during this time in 
the case of more sensitive paper irrespective of an acid soil condition, 
is on the surface a valid objection to the general use of litmus paper 
as a test for soil acidity. 

To obtain some information as to the color change in acid soils of 
blue litmus paper from different sources, samples of the paper were 
gotten from different departments in the experiment station. Litmus 
paper from five different sources was thus secured. Requests were 
also made of a number of druggists through the State for samples 
of their blue litmus paper and samples obtained from ten of them. 
In a number of cases the paper was not in stock. These samples 
were compared with the Kalbaum's paper by testing in a soil which 
had a limestone requirement of 2,912 pounds as determined by the 
lime-water method. 

The tests were made very much after the method used in labora- 
tories of the Misouri station by laying the litmus paper slips on a 
watch glass and forcing the soil after it had been made into a mud 
ball well up against them by means of a second watch glass. This 
leaves the soil and litmus paper between two watch glasses with the 
paper directly against one glass so that the color change can be ob- 
served at any time. 

Under these conditions, the Kalbaum's paper showed a distinct pink 
in 2 minutes, and the maximum pink color was reached in about 5 



1 82 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



minutes. Of the five experiment station papers, two were found to 
be about as sensitive as Kalbaum's, one required an hour to show a 
slight pink, and two at the close of two hours showed very little or 
no pink at all. 

The druggists' samples were compared with the Kalbaum's in the 
same way. In 5 minutes, as usual, the Kalbaum's paper gave a good 
pink. Of the druggists' papers, one gave a good pink and 3 others a 
slight pink. In 30 minutes, slight pink appeared in 3 more. At the 
end of two hours, a good pink was present in 4 of the druggists' papers, 
a fair amount of pink in 3, and a slight pink in one. In paper from 2 
of these samples no pink color appeared at all. 

W hen these druggists' papers together with the Kalbaum's were 
placed in one thousandth normal hydrochloric acid solution, the fol- 
lowing color changes appeared. The maximum amount of pink ap- 
peared in the Kalbaum's paper in a half minute, and in the druggists' 
papers as follows: In one, in 1.5 minutes; in one, in 5 minutes; in 
two, in 8 minutes; in two, in 9 minutes; in one, in 15 minutes, in one, 
in 30 minutes ; and in two, no pink color appeared at all. These 
latter two gave pink color in a strong acid solution. In the main, 
the sensitiveness of the papers as determined by the acid solution 
checked up closely with that observed in the soil. 

Tests in the soil were made with a few neutral litmus papers from 
different sources, but the color change is not distinct enough to make 
the use of this paper advisable. 

Tests were also made with the Kalbaum's blue litmus paper in a 
few soils of limestone origin which still contained limestone as shown 
by vigorous effervescence with hydrochloric acid, to determine whether 
a pink color would appear under these conditions. In a heavy sub- 
soil a slight pink appeared at the end of 30 minutes; however, it was 
not permanent. In a lighter surface soil the paper always remained 
blue. There is no doubt, however, of the tendency for sensitive blue 
litmus paper to develop a pink color in soils of high colloidal content 

in the absence of an acid condition. 

In pail ai a result of this limited study, it is thought inadvisable 
to recommend the blue litmus paper test to farmers as a means of 
•' ' • on Mil acidity. In the hands of an operator who is familiar 
with the paper he is using and knows what color changes to expect 

und< r various conditions, the test is a good qualitative one for soil 
audits and m addition givet ome information in a quantitative way, 

hut tor general USe it can not be considered reliable and may give 
n Wilts whieh are entirely misleading. 



biggar: maize seed preparation. 



PRIMITIVE METHODS OF MAIZE SEED PREPARATION. 1 

H. Howard Biggar. 

The word " corn " in the Indian language has many forms. In the 
Sioux it is "wagamaza," in the Omaha it is " wahaba," in the Gros 
Ventre it is " holiati," in the Mandan it is " khati," while the Arikara 
calls it " nicissee." The word is one of the most important in the 
tribal vocabulary, since corn for generations was the main and often 
the only food plant. 

In an investigation covering 15 Indian reservations in Minnesota, 
North Dakota, South Dakota, Nebraska, Montana and Manitoba, the 
writer was much impressed by the agricultural practices of the In- 
dians in connection with their corn production. None of these prac- 
tices are of more interest than the preparation of seed. 

The Indian designated time by referring to natural phenomena. 
Seed was prepared and corn planted when the wild turnips began to 
bloom, when the grass began to become green, when plums, wild 
grapes, or juneberries began to blossom, when the leaves began to un- 
curl, or when the first prairie flowers began to bloom. Superstition 
and suggestive magic played an important part in seed preparation. 
Red Bear, an Arikara of the Fort Berthold reservation, informed me 
that the oldest woman of each family was usually intrusted with 
this work and that it was partly a secret process and almost a sacred 
one. It was Red Bear's opinion that since the old methods had been 
discontinued the cornfields of the Indians produced lower yields and 
the plants were more susceptible to insects and to plant diseases. 

Various standards were used as the basis of seed selection. Many 
of the Indians have told me that moldy cobs on ears in the fall were 
very undesirable. For the most part, well-filled tips were sought as 
well as straight rows of kernels. The tip and the butt kernels were 
discarded and the middle kernels used for seed, the explanation being 
that these were better producers. Seed ears were braided together 
by the husks every fall, the braids being about 5 feet long. 

Indian informants on the Crow Creek, Lower Brule, Rosebud, 
Yankton, Standing Rock, Fort Berthold, Fort Totten, and Red Lake 
reservations, representing tribes of the Sioux, Gros Ventre, Arikara, 

1 Contribution from the Office of Corn Investigations, Bureau of Plant In- 
dustry, U. S. Department of Agriculture, Washington. D. C. Received for 
publication February 2, 1918. 



I 84 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

and Chippewa Indians, stated that it was always the custom to sprout 
the corn kernels previous to planting. The sprouting had a two- fold 
object, (i) to insure the plants coming up more quickly and (2) to 
insure a good stand of plants, since kernels not sprouting were at 
once discarded. 

In the Red Lake reservation country of northern Minnesota, the 
Chippewa Indians had their fields bordering the lake and entirely 
around it. The lake country was densely wooded and hence use was 
made of moss in sprouting. Previous to planting, a box was filled 
with moss and kernels of corn placed in the moss. The whole was 
soaked for a time and then set in a warm place until the kernels 
sprouted. Dead kernels were thrown out and the sprouted ones were 
planted. Some tribes placed kernels in small sacks, soaked the sacks 
in water, and then hung them in a warm room to germinate. Others 
made willow baskets, filled them with kernels of corn, poured water 
through the corn, and then placed the baskets in a warm place. 

Red Horse of the Yankton Reservation was the first to tell the 
writer of the use of the ground plum (Astragalus caryocarpus) in 
connection with the sprouting of seed corn. Later, members of other 
tribes corroborated his story. It is a custom among some tribes to 
soak the fruits of the ground plum in the same water in which the 
seed corn was soaked. The ground plum is the well-known prairie 
legume with a habitat ranging from Saskatchewan to Texas. Dr. M. 
R. Cilmore. State historian of North Dakota, believes that as the 
ground plum is prolific, bearing many fruits, the Indians thought its 
use would make the corn more productive. This explanation is in 
line with Indian beliefs. The Omahas in Nebraska placed the shell 
of a turtle in the water with the corn, believing that, as turtles eat 
insects, the use of the shell would aid in keeping the corn plants free 
from insect damage. 

W hile the write r was visiting the Crow Creek Indian Reservation 
in South Dakota, Medicine ( row. Kill Dead, and Seeking Land gave 
tlx- first information regarding what might almost be called a primi- 
tive ragdoll seed tester. Subsequently William Bean of the Yankton 
I-'- ■ r vation in South Dakota and Malokikla and Little Bull of the 

Port Totter Reservation in North Dakota confirmed their statements. 

The main pari of this primitive tester was the stein of the slender 
nettle, Urtico gracilis, The leaves of this nettle are sparingly armed 
with Itinging hairs. After describing the plant, Medicine Crow took 
me into the near by woods and showed me a clump of them. Large 

patches were also ' ' 1 1 1 1 1 1 1 1 1 vicinity of Devil's Lake, N ' . Dak ., on the 



biggar: maize seed preparation. 



185 



Fort Totten reservation. The Sioux Indian name for this nettle is 
" asbehu " or itch weed. 

The slender nettle was used in the following manner. When the 
time for planting corn was at hand, quantities of the nettle were 
gathered. They were then piled up in a sort of mat and on this 
mat the kernels were placed. The mat of nettles was then rolled up 
so that it made a cylindrical bundle with the corn kernels on the 
inside. The bundle was tied round with strings cut from buffalo hide 
and then immersed in water. After soaking for a day or more, the 
bundle was removed, wrapped in a buffalo skin or some other cov- 
ering, and kept warm. In a few days the kernels sprouted and when 
the sprouts were a quarter of an inch or more long they were planted. 
Kernels not sprouting or showing swollen germs were at once dis- 
carded. The slender nettle was used for this purpose instead of some 
other prairie plant because it was the first plant to reach any consid- 
erable height by corn-planting time. Further, the fact that the plant 
was protected with stinging hairs gave the Indian the superstitious 
idea that corn germinated with it would be protected from plant ene- 
mies during the growing season. 

At the South Dakota State Fair at Huron in September, 1917, the 
Crow Creek Indian Agency exhibit included a nettle tester in which 
corn was being germinated. These testers have been used by the 
Indians for at least seventy-five years. 

The principal corn grown by the Indians of the Middle West was 
the soft or flour corn, Zea amylacea. This type absorbs water more 
readily than the dent or flint types, as shown by the following experi- 
ment. One hundred grams of each variety were placed between wet 
bleached muslin and reweighed after 22 hours. The blue flour corn 
absorbed 18.0 percent of water; Reid yellow dent absorbed 13.5 per- 
cent; and U. S. Selection No. 193, a flint corn, only 6.6 percent. 

THE TIME AT WHICH COTTON USES THE MOST MOISTURE. 1 

C. K. aIcClelland. 

It has long been recognized that cotton can be produced with less 
water than is required for crops of corn and oats. Not only is this 
true, but apparently there is a great difference in the stage of growth 
when these crops require the maximum amounts of water. A dry 

1 Contribution from the Georgia Agricultural Experiment Station, Experi- 
ment, Ga. Presented by the author at the second annual meeting of the Asso- 
ciation of Southern Agricultural Workers, New Orleans, La., in January, 1917- 
Received for publication February 10. 1918. 



I 86 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



time occurring when corn or oats are filling and ripening notice- 
ably reduces the yield, but with cotton a dry time during the maturing 
(opening) of the crop is to be desired. Probably, however, there is 
an analogy between the processes of growth which are taking place 
in the cotton and in the grains during the periods when each is using 
iarge amounts of water ; that is, the blooming and boll-filling period of 
cotton rather than the opening should be considered the same period 
of growth as the blooming and head-filling or ear-filling periods of 
the grains. At this period of growth a dry time is equally as disas- 
trous with cotton as with the grains, the effects being shown largely 
in the shedding of the squares or newly set bolls. 



Table i. — Liters of water applied weekly from July to September, inclusive, 
to individual cotton plants grown in potometers in 1915, with the per- 
centage of saturation at which each was maintained. 



Plant No. 


Saturation, 
percent. 


First period. 


Second period. 


Third period. 


Week in July. 


4th 
week 
July 


Week in August. 


1st 
week 
Sept. 


Week in Sept. 


1st. 


2d. 


3d. 


1st. 


2d. 


3d. 


4 th. 


2d. 


3d. 


17 
24 
6 
18 
29 
14 
23 
22 

7 
3<J 


60 

73 
75 
75 
90 
90 

30-60-90 
90-30-60 
90-30-60 


3-5 
3-5 

1.8 
1.8 
1.8 
3-5 
3-5 


3-5 
5-3 
7.0 
1.8 
21.0 
3-5 
1.8 
17-5 
15-5 


3-5 
28.0 
15-7 
14.0 
14.0 
22.8 
17-5 
14.0 
15-7 
24-5 


12.8 
28.7 
19.9 

23- 4 
13.2 
39-8 

24- 3 
18.4 

6.6 
6.6 


6.6 

35-4 
26.5 
44.2 
22.1 
50.8 
42.0 

15-5 
13.2 
28.7 


19.6 
24.O 
37-6 
26.5 
28.7 
28.7 
28.7 
15-5 
2.2 

24-3 


19.9 
24-3 
35-4 
31.0 
26.5 
28.7 
28.7 
31-0 
22.1 
24-3 


31.0 

42.0 

33.2 

37.6 

35-4 
28.7 
26.5 
22.1 
11. 
15-5 


15-5 
II. 
19.9 
17.7 
II. 

6.6 
6.6 
13.2 
6.6 
4.4 


26.5 
26.5 
26.5 
19.9 
26.5 
28.7 
28,7 
33-2 
15-5 
31-0 


19.9 
22.1 
13.2 

15-5 

22.1 

15-5 

4.4 
24.3 

8.8 
4.4 


Totals . . . 





[9.4 


77.x 


169.7 


193-7 


285.0 


232.8 


271.9 


293.0 


112. 5 


263.0 


150.2 



Blooming records taken in different years may show a seasonal 
variation and naturally there will be some variation between varieties 
and likewise between individuals. In general, however, records of 
blooming will give a fairly good index to the time of year when 
ill- filling process begins. The time when the blooms are coming out 
in"- 1 rapidls and the time when the plants are using the most mois- 
ture • • in to be Correlated. In experiments to determine the water 
requirement! of cotton conducted at the (Georgia station in [915 and 

1916, plant! were grown in potometers and the quantity of water 

applied .it different periods recorded. Table i shows the liters of 
water ap plied to the individual plants by weeks during July, August, 
and September, The number of blooms appearing each week on each 
individual and the number of blooms appearing on plants under field 
conditions each week arc reeorded in Table 2. 



m'clelland: use of moisture by cotton. 



187 



Table 2. — Number of flowers opening weekly on individual cotton plants in 
potometers in iQij, with the number on 18 plants in the field. 



Plant No. 



Total 
number 
of blooms. 



84 
67 
67 
47 
35 



Week in July. 



Week in August. 



First week 



Third. Fourth. First. Second. Third. Fourth. 



24 
18 
17 
9 
1 



20 
13 

9 
11 

I 



Total 

18 field plants. 



300 
327 



36 
129 



48 



76 
80 



6 9 



54 



Table 2 shows that the plants in the potometers began blooming the 
third week in July, increasing until the second week in August, and 
then decreasing. The record of 18 plants in the field showed that 
the greater number of blooms appeared there during the last week in 
July. This difference is due primarily to the fact that the plants in 
the potometers were from seed planted rather later than those in the 
field and even in most cases from replantings. The replanting was 
due to the destruction of the first plants by girdling with plastic clay 
in attempts to prevent evaporation and the penetration of rain water 
into the cans. The cotton plant is very tender in its earlier stages 
and is easily damaged by such treatment. If the supply of moisture 
remains constant, blooming may be delayed. This statement is per- 
haps better made in the reverse order, that is, the checking of the 
supply of moisture induces earlier blooming. In the fields the supply 
was checked, there being no heavy rains from July 4 to August 16. 
In the cans, the supply was constant except as noted. With plants 
22, 7, and 30 the saturation was changed as shown in the table during 
each period. Plant No. 7, where the moisture was reduced from 90 
percent to 30 percent, produced the maximum number of blooms one 
week earlier than the average. Plant No. 22 bloomed later than the 
average due at least in part to the accidental topping of the plant. 

Most of the plants w r ere grown in soils where the saturation was kept 
constant or nearly so, semiweekly w r eighings being made to show T the loss 
of water and the depreciation below the desired content, after which 
water was added to bring the saturation to the desired point or a little 
above. Where the saturation was varied, the change was made when 
the squares began to form and when the blooming period was nearly 
over. It will be noted that the plants used large amounts of water 
during the entire second period and well into the third. On account 
of the breaking of the balances with which the weighings were made, 



I 88 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



no records were obtained after the third week in September. The 
second period covered nearly the entire blooming period of the plants. 

Records similar to those shown in Tables I and 2 for 191 5 are pre- 
sented in Tables 3 and 4 for the experiments conducted in 1916. 

Table 3. — Liters of "water applied weekly from July <?5 to October 7, to indi- 
vidual cotton plants grown in potomcters in 1916, with the percentage 
of saturation at which each was maintained. 



Week ending 



Plant No. 


Saturation, 
percent. 


July 

25- 


Aug. 8. 


Aug. 14. 


Aug. 22. 


Aug. 28. 


Sept. 
4 


Sept. 
12. 


Sept 
19. 


Sept. 

25. 


Oct. 7. 


4 


45 


2.25 


9.00 


13-50 


II.25 


13-50 


6-75 


II.25 




6-75 


4-50 


11 


45 




6-75 


11.25 


II.25 


13-50 


6-75 


II.25 




9.00 


4-50 


17 


45 




2.25 


9.00 


9.00 


13-50 


2.25 


II.25 




9.00 


4-50 


36 


45 




4-50 


13-50 


9.00 


13-50 


4-50 


II.25 


2.25 


9.00 


4-50 


12 


60 


2.25 


13-50 


18.OO 


H.25 


15-75 


11.25 


4-50 


13-50 


II.25 


4-50 


18 


60 




13-50 


13-50 


13-50 


9.00 


9.00 


13.50 


6-75 


6-75 


4-50 


4i 


60 




13.50 


13-50 


15-75 


11.25 


9.00 


15.75 




6-75 


4-.SO 


43 


60 


2.25 


H.25 


9.00 


i3-5o 


11.25 


9.00 


13-50 


2.25 


6.75 


4-50 


30 


75 


9.00 


15-75 


20.25 


9.00 


18.00 


9.00 


9.00 




6-75 


4-50 


3i 


75 




4-50 


II.25 


9.00 


9.00 


6-75 


2.25 


2.25 


6-75 


4-50 


33 


75 


9.00 


9.00 


24.75 


15-75 


11.25 


11.25 


6-75 




6-75 


4-50 


Totals . . . 




24-75 


103-50 


I57-50 


128.25 


139-50 


85.50 


49.00 


27.00 


85.50 


40.50 



Table 4. — Number of flowers opening weekly on individual cotton plants grown 
in potomcters in 1916, with the number on to plants in the field. 





Week ending 


Plant number. 


















July 25. 


Aug. 8. 


Aug. 14. 


Aug. 22. 


Aug. 28. 


Sept. 4. 


Sept. 12. 


4 




I 


4 


10 


8 


5 


5 


1 1 








4 


9 


1 1 


16 


17 






3 


8 


9 


5 


4 


36 




3 


5 


10 


8 


9 




12 




3 


8 


15 


14 


9 




18 






'1 


8 


1 1 


7 


4 


4i 




2 


4 


10 


7 


7 


6 


43 




2 


4 


8 


8 


9 


2 


3o 




7 


5 


12 


6 


3 




31 






2 


5 


2 


9 


6 


33 




15 


8 


16 


3 










33 


44 


105 


85 


74 


42 


Total on 10 field plant* . . 


53 


44 


85 


146 


58 


37 


8 



The water record in \<)\(> was carried a little later than in T915, 

and ihowed a eon idnablr demand by the plant well into the fall 
months I lie greatest amount of water was applied to the plants 
during tli' 1 four weeks of August, the highest week being the second. 
On account of the h<av\ rainfall on July 15 the field plants did not 



m'clelland: use of moisture by cotton. 189 

bloom earlier than the plants in the cans ; the maximum blooming 
period for both was during the third week in August. Except for 
No. 11 which was accidentally topped, the plants in the cans were 
quite regular in their period of blooming. 

Most farmers are not satisfied if they have not laid their cotton by 
so that they are free to go fishing by July 4. If the results here pre- 
sented are of any value, they indicate that later cultivation than is 
usually given would be of benefit in conserving soil moisture for the 
use of the plants during the hot summer weather. 

AGRONOMIC AFFAIRS. 

OFFICIAL CHANGES. 

Because of a change in his official duties and his removal from 
Washington incident thereto, Mr. P. V. Cardon has resigned as Sec- 
retary-Treasurer of the Society, effective March 15. Mr. Lyman 
Carrier, of the Office of Forage-Crop Investigations, U. S. Depart- 
ment of Agriculture, and a former Treasurer of the Society, has been 
appointed acting Secretary-Treasurer by President Lyon, and all cor- 
respondence regarding memberships and dues should be addressed to 
him at Washington. 

Attention is also called to the changes in the standing committees 
of the Society, as shown on the back cover page of this number. 
On the Committee on Soil Classification and Mapping J. G. Mosier 
and C. A. Mooers have succeeded G. N. Coffey and L. J. Briggs. 
President Lyon has removed himself as chairman of the Committee 
on Standardization of Field Experiments and has appointed A. T. 
Wiancko, a former member of the committee, in his stead; new mem- 
bers on this committee are F. S. Harris and S. C. Salmon, the latter 
succeeding W. M. Jardine. No changes were made in the Committee 
on Terminology. On the Committee on Varietal Nomenclature, W. 
C Etheridge has succeeded C. G Williams. The whole-hearted co- 
operation of the entire membership with these new officers and com- 
mitteemen is earnestly solicited. 

ANNUAL DUES FOR 1918. 

Under the by-laws of the Society, the Journal is not to be sent to 
those who have not paid their dues by April 1. Because of a change 
in the Secretary-Treasurership and other conditions which have some- 



I9O JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



what delayed collections this year, the April number is being sent to 
all members whose dues for 191 7 are paid, as well as all new mem- 
bers for 1918. Before the May number is mailed, however, the 
payment of dues will be checked and those who are in arrears for 
1918 will be taken off the mailing list. If you have not already 
paid your dues for 191 8 and wish to continue to receive the Journal 
promptly you should remit the amount, $2.50, promptly to the new 
Secretary-Treasurer, Mr. Lyman Carrier, U. S. Department of Agri- 
culture, Washington, D. C. 

MEMBERSHIP CHANGES. 

The membership reported in the March issue was 638. Since that 
time 9 new members have been added, 2 have been reinstated, and 1 
has resigned, a net gain of 10 members and a present membership 
of 648. The names and addresses of the new members and of the 
two reinstated, the name of the one resigned, and such changes of 
address as have been noted since the last issue, are as follows : 

New Members. 

Clevexger, C. B., College of Agriculture, Madison, Wis. 

Gillis, M. C, 401 E. Douglas St., Bloomington, 111. 

IBERICO, Juan R.. Yurimaguas, Loreto, Peru, S. A., via Para, Brazil. 

Love, Russell M., R. F. D. No. 2, Tarentum, Pa. 

RuNK, ('. R., 250 W. Tenth Ave., Columbus, Ohio. 

Smith, V. C, College of Agriculture, Columbus, Ohio. 

Vl \< h, C L., College of Agriculture, Athens, Ga. 

Water, E. J.. 91 West nth St., Columbus, Ohio. 

Young, Philip H., Kans. State Agr. College, Manhattan, Kans. 

Members Reinstated. 

McCall, M. A., Lind, Wash. 

RoBT., E liattery, Can. Anti-Aircraft B.E.F., France, via P. M., New 
York, N. Y. 

M 1 m beb Resigned. 

Si i»i»ath, Robert O. 

Changes of Address. 
Bi\H»Kt), K. K„ Dadeville, Ala. 

Budsos, k. Pagb, Experiment Station, WaterviUe, Wash, 

Hkyant. Kay. Stillwater, Okla. 

Buksth, Dan M m R. F. D. No. 3, Chanute, Kans. 

J'.i smv. A. I,. Plankinton, S. I). 

( 1 . f. S., Starkvillc, Miss. 

( iiai'Man, Jamks I*:., 1812 Lindm Ave, Baltimore, Md. 



AGRONOMIC AFFAIRS. 



Deatrick, Eugene, Mont Alto, Pa. 

Douglas, J. P., 402 E. Chalmers St., Champaign, 111. 

Kemp, Arnold M., Fairmount, Ind. 

Kennard, F. L., Colfax, Wash. 

Lechner, H. J., Court House, Astoria, Ore. 

Morrison, J. D., Elbon, S. Dak. 

Nash, C. W., Morris, Minn. 

Thomas, Melvin, Charleston, 111. 

Waller, Allen G., 2028 F St., NW., Washington, D. C. 
Walster, H. L., 5520 Blackstone Ave., Chicago, 111. 

Mail for the following members has been returned unclaimed. Any 
one knowing the present address of any of these members will 
confer a favor on the Secretary-Treasurer by reporting the infor- 
mation to him. 

Bliss, S. W., Freeman, Ray, Hurst, J. B. 

Boardman, W. C, 



ROLL OF HONOR. 

In the February Journal a list of those members of the Society who 
were known to be serving their country in its military forces was pub- 
lished. Since that time, a number of names have been added to the 
list. The editor will appreciate the favor if those who know of other 
members of the Society than those noted below who should be added 
to this list will send their names to him. 



Samuel D. Gray, 
E. E. Graham, 
M. B. Gilbert, 
A. D. Ellison, 
James E. Chapman, 
H. R. Cates, 
A. M. Brunson, 



B. B. Holland, 
O. F. Jensen, 

C. H. Karlstad, 
P. H. Kime, 
Leroy Moomaw, 
j. a. purington, 
j. v. quigley, 



Geo. T. Ratliffe, 
L. C. Raymond, 
Phil. E. Richards, 
f. j. schneiderhan, 
w. r. schoonover, 
Herschel Scott, 
Paul Tabor. 



NOTES AND NEWS. 

G. M. Garren is now assistant agronomist in plant breeding at the 
North Carolina station. 

E. J. Holben has been appointed assistant in experimental agron- 
omy at the Pennsylvania station. 

R. R. Hudelson, assistant professor of soils in the University of 
Missouri, is now a first lieutenant of artillery and E. M. McDonald, 
assistant professor of farm crops in the same institution, is a second 
lieutenant of infantry in the National Army. 



1 9 2 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



W. M. Jardine, president of the American Society of Agronomy 
in 1 91 7 and for the past several years dean and director of the Kansas 
college and station, has been elected president of that institution, suc- 
ceeding H. J.. "Waters, who, as previously noted, is now managing 
editor of the Kansas City Weekly Star. 

H. G. Knight, for the past several years dean and director of the 
Wyoming college and station, has been elected to a similar position 
in Oklahoma and assumed his new duties February 1. He was suc- 
ceeded in Wyoming by A. D. Faville. 

F. G. Merkle, who has been assistant in the department of agron- 
omy in the Massachusetts Agricultural College during the past year, 
has been made an instructor in the same department. 

M. F. Miller, professor of soils in the University of Missouri, has 
been made assistant dean of the college and assistant director of the 
station in addition to his other duties. 

J. A. Purington, of the Massachusetts college, enlisted in the 
U. S. army last December. He was on the Tuscania when she was 
sunk by a submarine in February, but was fortunate enough to be 
numbered among the survivors and at last reports was somewhere in 
England. 

H. N. Vinall has been at Wichita, Kans., for the past two months, 
supervising purchases of grain sorghum and other seeds for sale 
to farmers at cost in the drought-stricken regions of Oklahoma and 
Texas where there was a general failure last year. This work is 
being done for the Seed Stocks Committee of the U. S. Department 
of Agriculture. 

Meeting of the Ohio Section. 

The Columbus (Ohio) Section of the American Society of Agron- 
omy held it- second annual meeting on Wednesday afternoon, Janu- 
ary 30, [918, in Townsend I [all, ( >hio State University. At this meet- 
ing Mr. J. W. Ames, chief chemist of the Ohio Agricultural Exper- 
iment Station. di8CU9«ed " Sulfofication in Relation lo Nitrogen Trans- 
portation in Soils.'' Dr. |. I\ Lyman, Department of Agricultural 
Chemistry and Soils of the Ohio State University, discussed "The 
Food Problem and the War." About 30 men were in attendance at 
the meeting, and look part in the discussion of these two subjects. 
The meeting was very profitable and enjoyable. The officers elected 
for the coming year wen-: President. Kirman K. Hear; and Secretary- 
Treasurer, W allace K. I Linger. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. io. May, 1918. No. 5. 



THE EFFECT OF SODIUM NITRATE APPLIED AT DIFFERENT 
STAGES OF GROWTH ON YIELD, COMPOSITION, AND 
QUALITY OF WHEAT — 2. 1 

J. Davidson and J. A. Le Clerc. 
Introduction. 

In a previous paper 2 were reported the effects of sodium nitrate 
applied at different stages of growth on the yield of crop, percentage 
of yellowberry, and protein content of the grain. It was shown that 
when applied at the first stage, sodium nitrate increased very consid- 
erably the yield of the crop, that when applied at the second stage 
it increased the protein content and " flintiness " of the grain, and 
that when applied at the third stage it did not have any effect either 
on the yield of the crop or on the composition and quality of the 
grain. Attention was called to the new method of plotting used in 
the experiment. The plots were laid out after the crop was up, each 
plot showing uniformity of plant growth. This made it possible to 
limit the size of the plot to one square rod with very satisfactory re- 
sults. It may be stated in passing that experiments carried out this 
year in Nebraska fully corroborate our observations made in our ex- 
periments conducted last year in Kentucky with reference to the par- 
ticular effect of nitrogen at the different stages of growth and with 

1 Contribution from the Plant Chemical Laboratory of the Bureau of Chem- 
istry, U. S. Department of Agriculture, Washington, D. C. Presented at the 
tenth annual meeting of the American Society of Agronomy, Washington, 
D. C, November 13, 1917. 

2 Davidson, J., and Le Clerc, J. A. The effect of sodium nitrate applied at 
different stages of growth on the yield, composition and quality of wheat. In 
Jour. Amer. Soc. Agron., 9: 145-154. 1917. 

193 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



reference to adaptability of our new method of plotting for experi- 
ments of this kind. 

In this paper we want to present and discuss some additional data 
obtained in our last year's experiments, namely, the effect on the ash, 
potash, and phosphoric acid content of the grain, and on the nitrogen 
content of the straw. 

Ash, Phosphoric Acid and Potash. 

As seen from Table I, the various treatments in the experiment did 
not affect the ash, phosphoric acid, or the potash content of the grain 
in any distinct or concordant manner. The variations in the phos- 
phoric acid are on the whole very slight. Should these small varia- 



Table i. — Percentage of ash, phosphoric acid, and potash in the grain from 
plots to which nitrate of soda was applied in various quantities 
at various stages of growth. 





itt 

c 




Percentage of ash. 


Percentage of 
P 2 5 . 


Percentage of 
K 2 Q. 




c 




Water 




Water 




Water 




Fertilizer added at 
each application. 






applied. 


a 


applied. 


a 


applied. 


& 


"a 
a 

« 


Stages of growth. 






sd 














u a 


c 1 j 


V T3 


u c 




u • 


u a 


a ^ 


u-6 




o 




N -2 
H= 3 




■i ft 


SJ .2 

7Z 3 






N .2 
rr 3 




rt 

* P. 




o 




Fert 
in sol 


Fertil 
solid 


O 


Fert 
in sol 


Fertil 
solid 




ft 


Fert 
in sol 


Fertil 

solid 


O -j 


2 lbs. NaN0 3 




First 


I.90 


I.98 


2.05 


I.02 


•99 


1.00 


.542 


.524 


^538 












1.82 


1.86 


1-97 


•93 


.92 


.96 


.527 


.522 


•533 


Do 


I 


Second 


i-95 


1-95 


I.98 


I.03 


I.03 


1.04 


•55' 


•559 


•552 












1.90 


1.92 


1-99 


•99 


.98 


I. OOI .495 


•495 


.550 


Do. 




Third 


1.86 


1.89 


I.98 


•99 


I.03 


I.OI 


•539 


•555 


.540 














1-94 


1.88 


2.04 


1.00 


I. OO 


I.O3 


•533 .528 


•542 


i lb. NaN0 3 


2 


First and second 


i.9 


i-95 


1.82 


I. OO 


I.OI 


•94 


.552 


■555 


.502 








1.88 


i-75 


1.88 


.95 


.92 


.<>5 


•5<'7 


.49° 


.527 


Do. 


2 


Second and third 


1.97 


2.01 


2.17 


I.O^ 


I.OI 


I.05 


.548 .560 


.529 












1.91 


1.99 


2.08 


•97 


.95 


I.O4 


.487 .483 


•545 


Do. 


2 


First and third 


1.88 


i>5 


1.95 


.98 


1.00 


•95 


•544 .526 


•543 






% lb. NaNO, 




Pint, second 1 


1.97 


1.82 


1.83 


•97 


.90 


.91 


.508 


•478 


•493 


3 


2.09 


1-9' 


1.92 


I.02 


1.02 


I.OI 


.521 


• 555 


525 


2 lbs. NaN< >, 2 1 




and third J 


1.80 


1.92 


1.80 


•94 


.96 


•93 


.503 


•5°4 


.4^8 


i 


Pint 


i-57 


1.73 


.90 


.86 


.507 


.486 


lbs. KC1 ( 






1.91 




1.84 


•95 




1,02 


•533 




.526 


Do. 


i 


Second 


i.7« 




1.83 


.98 




•97 


.526 


•505 


Do. 




Third 


1.98 




1.87 


I.OI 




.98 


.504 




.500 


i 


1.77 




2.02 


.97 




•97 


•544 




•5'o 


2 Ibfc KC1 






1.93 




1.92 


.96 




I.03 


•5'8 




•543 


i 


1 i r- 1 


1. 86 
I.9.S 




2.06 

l No 


1.00 

1.03 




I.OI 

•97 


541 

.536 




•555 
.558 










Do. 


i 


Sf< r ,||«| 


- 7« 




1.93 


•97 


.98 


•534 




.528 


Do 




i bM 


1.9' 




2.07 


.99 




• 97 


• 5°9 




.502 




1.9 J 
1.97 




I.84 
I.89 


,98 
1.00 




.95 

1.00 


•549 
544 




• 5'3 

•539 


"Check 






2.04 


1 ,99 


1.00 


.9* 


•576"" 


.582 








2.02 


' 189 


99 


1.04 


•524 


.529 



DAVIDSON & LECLERC : EFFECT OF SODIUM NITRATE. 



195 



tions be taken into consideration there is possibly a very slight tend- 
ency toward a somewhat higher phosphoric acid content in those cases 
where the nitrogen was applied in the second stage. It would be 
useless, however, to speculate whether the nitrogen applied at the 
first stage caused a depression in the phosphoric acid or whether the 
nitrogen applied at the second stage stimulated the phosphoric as- 
similation. Headden 3 found a distinct depression in the phosphoric 
acid content when sodium nitrate was applied. More experimenta- 
tion is necessary in order to decide whether the results obtained by 
Headden are the rule, our results being the exception, or vice versa. 
It is likely that Headden s results are the rule, as there are certain 
theoretical considerations which would be in harmony with this sub- 
stitution of nitrogen for phosphorus. The case would be similar to 
the substitution of sodium for potassium. We will perhaps be able 
to throw some light on the question when we are ready to report on 
the results of our Nebraska experiments. With reference to the ash 
and the potash content there is no tendency toward any consistency 
whatever. 

Protein Content in Straw. 

The term protein with reference to straw represents the -total nitro- 
gen multiplied by 6.25. The protein content is given instead of the 
nitrogen content, both for the sake of uniformity and to accen- 
tuate the consistent variations in the nitrogen content. As seen from 
Table 2 the protein content in the straw follows the same tendency as 
the protein content in the grain. The straw of the plots which re- 
ceived the sodium nitrate at the second stage shows a distinctly in- 
creased protein content, the increase being proportional to the amount 
received. Of the plots which received their nitrogen in two stages 
those which received it in the first and second stage, and in the second 
and the third respectively show an appreciable increase of protein in 
the straw ; those, however, which received it in the first and the third 
stage do not show such «an increase. The same is true about the 
plots which received the full application in the first and third stages 
respectively. 

The application of potassium chloride depressed very distinctly the 
protein content of the straw. This observation is in full accordance 
with the results of Headden, who found that potassium chloride de- 
pressed the nitrogen in the plant but not in the grain. 4 From Table 

3 Headden, W. P. A study of Colorado wheat. Colo. Agr. Expt. Sta. 
Bui. 219. 
4 Loc. cit. 



I96 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



2 it will also be noted that the yield of straw was increased when- 
ever nitrates were added at the first sage. The application of KC1 
had no effect on the yield, irrespective at what stage the potash salt 
was applied. 

Table 2. — Percentage of protein in straw and weight of straw from plots to 
which nitrate of soda was applied in various quantities at various 
stages of growth. 



Fertilizer added at each 
application. 



Stages of growth. 



Percentage of protein in 
straw. 



Water applied. 



Ferti- 
lizer in 
solu- 
tion. 



Ferti- 
lizer in 
solid 
state. 



No 
water 
applied 



Weight of straw, lbs. 



Water applied. 



Ferti- 
lizer in 
solu- 



Ferti- 
lizer in 
solid 
state. 



No 
water 
applied. 



2 lbs. NfeNO, 

Do 

Do 

1 lb. NaNO, 

Do 

Do. 

% lb. NaNO s 

2 lbs. NaN0 3 + 2 ) 
lbs. KCI , 

Do. 

Do. 

2 lbs. KCI 

Do 

Do 

Check 



First 

Second 

Third 

First and second 

Second and third 

First and third 

First, second 

and third 
First 

Second 

Third 

First 

Second 

Third 



4.12 

4-74 
6.69 
6.50 

4- 55 
4.12 

5- i7 

4- 52 

5- 4i 

5- 83 
3 95 

3- 77 

4- 65 
4.82 
4-65 
3-95 

6- 37 
596 
4.21 
4.17 
324 
3 69 
3-5o 
3-82 
3.42 
3-<H 



4-47 

4- 39 
6.05 

5- 74 

4- 65 

4.43 
5.61 
4.78 
5.00 

5- 44 
4-47 
4.25 

509 
4-52 



395 
3.86 



4- 65 
4.91 
6.50 
6.62 
4 30 

3- 9i 

5- 44 
4.91 
6.25 
5-48 

4- 74 
4.68 

5- i7 
4-47 
3 77 

4- 47 

5- 44 
5-92 
3.68 
3 .86 
3-24 
3-99 
3-77 
3-64 
3-32 
3-59 

4.04 
4.21 



22. 1 
20 2 
9-5 
9-4 
10.3 
8.4 
16.3 
17-5 
1 1 .0 

7-4 
17. 1 
16. 1 
12.7 

13-4 
20.0 
24.6 
9.0 
12.0 

9-3 
1 1 . 2 

8.8 
10.8 
10.5 
11. 6 

9.1 

9-9 



22.8 
17.8 
10 8 
9-4 
9-5 
9-i 
15-9 
15-7 
9-3 
8.6 

15-5 
15-7 
12.6 

13-9 



20.4 
19.7 
9.8 
7-4 
9-8 
9-i 
16.3 
16.8 

9-3 
8.2 

15-7 
15-9 
14.9 
16.4 
19.4 
21.8 

9.1 
10.6 

7.8 

9- 
1 1 . 



12.4 
9 1 



9.2 
10.2 
9.6 

8.4 
10.0 

8.5 
10.9 



These results bear out further our conclusion that within the limits 
Of OUf experiment it is the presence of nitrogen in the soil at the 
I lage which is responsible for the increase of the protein con- 
tent in the Traill ;m '' straw and for prevention of yellowherry in the 
grain. Eieadden's 1 work can be interpreted to agree with the results 
obtained in our experiment. Ileadden applied his nitrogen at a 

1 Hcaddcn, W. P. Yellow berry in wheat, its cause and prevention. Colo. 
Af(r. Kxpt. Sta. Bui. 205. 



DAVIDSON & LECLERC: EFFECT OF SODIUM NITRATE. 197 



period which corresponded to our first stage and obtained a flinty 
grain high in protein. But the crop in his experiments did not re- 
spond in yield to the application of the nitrate, as nitrogen is not the 
limiting factor in the soils of Fort Collins. The added nitrate was 
not used up by the plant during the first stage of growth. The local 
condition further excludes the possibility of the removal of the 
nitrates by drainage. The added sodium nitrate consequently was 
present in the soil at the time which corresponds to our second stage 
of growth. It is the presence of the nitrate at this stage which, in 
our opinion, prevented yellowberry and produced a high protein 
content in the experiments of Headden. 

Headden 6 comes further to the conclusion, on the basis of his ex- 
periments, that flintiness and high protein content are a function of 
the soil and not of climate. According to the conclusion of Le Clerc 
and his associates 7 as a result of three years' experiments, climate 
and not soil played the predominant part in influencing the composi- 
tion and flintiness of wheat. In reality, however, the results ob- 
tained by Headden are not necessarily at variance with the conclu- 
sions reached in the Laboratory of Plant Chemistry. Under ordinary 
field conditions the crops depend for their nitrogen upon nitrifica- 
tion which is favored or hindered to a very considerable extent by 
climatic conditions. If Headden obtained the same results in a year 
which was very unfavorable with reference to rainfall as were ob- 
tained in a normal year it was probably because in his experiments 
the crop did not depend upon nitrates produced in the soil during 
that year, as he added ready-made sodium nitrate. 

Summary. 

1. No conclusion can be drawn from our experiments with ref- 
erence to the effect of sodium nitrate on the ash, phosphoric acid, 
and potash content of the grain. 

2. The protein content in the straw showed the same tendencies 
as the protein content in the grain, i. e., there was an increased protein 
content as a result of the nitrate applied at the second stage. 

3. An increase in the yield of straw as well as of grain was noted 
whenever nitrates were applied at the first stage. 

6 Headden, W. P. A study of Colorado wheat. Colo. Agr. Expt. Sta. 
Bui. 219. 

7 Le Clerc, J. A., and Leavitt, S. Trilocal experiments on the influence of 
environment on the composition of wheat. U. S. Dept. Agr., Bur. Chem. 
Bui. 128. 

Le Clerc, J. A., and Yoder, P. A. Environmental influences on the physical 
and chemical characteristics of wheat. In Jour. Agr. Research, 1 : 275. 1914- 



I98 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



4. Potassium chloride depressed the protein content in the straw. 

5. Headden's work can be interpreted to agree with the results of 

our experiments. 

6. The results of Headden's experiments as well as our own are 
not opposed to the conclusion of Le Clerc and his associates that 
flintiness and high protein content are a function of climate. 

Erratum. — In the first paper by these authors, published in Jour. Amer. Soc. 
Agron., vol. 9, no. 4, on p. 150, Table 1, read "Yield of crop (grain and straw), 
pounds." instead of rt Yield of grain, pounds." 



THE DETERMINATION OF MOISTURE IN SOILS. 1 

B. S. Davisson and G. K. Sivaslian. 
Introduction. 

The method commonly employed for determining moisture in soils 
is that of direct drying in an air oven heated by various means, as 
steam, gas, or electricity. A desire to overcome the objectionable 
features of these methods and to obtain an absolutely dry basis in- 
stigated the work reported in this paper. The objections which 
should be overcome are the long time necessary for removing the 
moisture, the contamination of the sample by gases from combustion, 
the oxidation of the organic material due to the long heating at a 
temperature of 105 to no° C. in air ovens, and the incomplete re- 
moval of all the moisture from the sample. 

Historical. 

In [898 Tryller 2 constructed a drying oven in which the heat from 
the small ring burners passed through a series of spaces in the walls 
and out al the top. The idea of this construction is to prevent contact 
between the ^oils and gases resulting from combustion. The author 
considered the results obtained by this oven to be sufficiently con- 
cordant for accurate work. I lowever. drying for fourteen days at 
105 C. and cooling and weighing daily did not yield absolute nor 
constant results. Often the sample shows an increase in weight after 

1 Contribution from the Laboratory of Soil Technology, Ohio Agricultural 
Experiment Station, Woostcr, Ohio. Received for publication February 
23. 1918. 

Tryller, . Uber die Beittmmung der Trokensubstanz irn Torf, In 

Lawlw. Versuchs., 40: 145. 1K0K. 



DAVISSON & SIVASLIAN '. MOISTURE IN SOILS. 



it has decreased considerably. The author maintains that drying with 
his oven is essentially as satisfactory as drying in a vacuum desiccator. 

Puchner 3 discusses Tryller's drying oven and admits that it pre- 
vents, to a certain extent, the action of gaseous products of combus- 
tion. He holds that the irregularity of the results obtained by Tryller 
are due to the unequal heating of different parts of the oven and the 
condensation of mositure on the glass and the soil, due to the high 
hygroscopicity, while transferring the sample to the desiccator. He 
then carried out some experiments in which he covered the dishes 
in the oven and left them covered when cooling in the desiccator. 
Better results were obtained in this way but they were not absolute. 
Puchner concludes that drying of a soil at 105 C. in such an oven 
does not give the actual moisture content and that soil analysis based 
upon such moisture determinations cannot be entirely correct. 

Mitscherlich 4 gave some time to accurate moisture determinations in 
soils and finally recommended drying in a vacuum over phosphorus 
pentoxide at ioo° C. He used individual vacuum desiccators with 
phosphorus pentoxide for each sample. The desiccator after evacua- 
tion was placed in an air oven at the desired temperature. Because of 
the thickness of the walls of the desiccator it was found that the tem- 
perature of ioo° C. within the desiccator could be more easily ob- 
tained by suspending the vessel in an oven and allowing live steam to 
come in contact with the walls of the desiccator. Mitscherlich em- 
ployed wheat starch for studying the drying process because it re- 
mained unchanged during the drying. He found that increasing the 
temperature lessened greatly the time necessary for drying the sample. 
Four hours' drying was found sufficient to remove all the moisture 
from the starch or soil sample at ioo° C, thus showing that the 
method gives absolute dryness, the dry substance remaining constant 
during five to nine hours' further drying. The author calls attention 
to the error introduced by handling the drying dishes with the bare 
hands. 

Konig 5 used an electrically heated vacuum desiccator in which he 
maintained a temperature of ioo° C. and used phosphorus pentoxide 
as the desiccant. The weighings were repeated to constant weight. 
A comparison of the results obtained by this method and by direct 

3 Puchner, H. The determination of dry matter in soil samples. In Landw. 
Versuchs., 55 : 309. 1901. 

4 Mitscherlich, A. Zur Methodik der Bestimmung der Benutzungswarme 
des Akerbodens. In Landw. Jahrb. 31 : 578. 1902. 

5 Konig, J. Relations between the properties of soils and the assimilation of 
foods by plants. In Landw. Versuchs., 66: 415. 1907. 



200 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



drying at 105 to no° C. does not show any very marked difference 
in the amount of moisture obtained upon different soil samples. 

Although considerable attention has been given to the subject of 
moisture determinations in soils, the positive evidence of the su- 
periority of any method is not conclusively established. It must be 
admitted that direct drying in an air oven does not give absolutely 
trustworthy results. The drying in a vacuum over phosphorus pen- 
toxide at ioo° C. appears to be superior to drying in an air oven. 
The time necessary for a moisture determination is greatly lessened 
by the vacuum method. 

A comparison of the methods reported seemed necessary to estab- 
lish fully the method which gives the most reliable results. A com- 
parison of direct drying in gas and electrically heated ovens and in a 
vacuum over phosphorus pentoxide at 105 C. constitutes the experi- 
mental part of this paper. 

Experimental. 

The gas oven used in this work is the ordinary copper oven pro- 
vided with a jacket for a liquid bath. Oil was used in this oven and 
by careful manipulation the temperature was maintained very con- 
stant at 105 C. No precaution was taken to prevent the gas fumes 
from coming in contact with the samples as they were being dried. 
The oven was used as it is generally in laboratories for drying. The 
electrically heated oven is one of the common types on the market 
having a control thermostat. The temperature was kept constant at 
105 C. at the shelf on which the samples were placed. 

In the first work of drying in a vacuum at 105 C. over phosphorus 
pentoxide, desiccators were used, but it was found practically impos- 
sible to maintain a vacuum in other than a glass desiccator and the 
glass desiccators which were obtainable would not withstand the tem- 
perature. An electrically heated vacuum oven was obtained and was 
found to be very satisfactory. Trouble was experienced in getting a 
tight seal between the door and the chamber without the use of some 

kind of sealing material. For this purpose a mixture was prepared 

by mixing finely divided graphite and melted beeswax until the mix- 
lure became very thick and pasty. After cooling, this material was 
cut into pencils. A pencil was then drawn around the face of the 
warm door until a good coating of the contact face was obtained. 
1 fpdn ' lo ing ili< door this graphite sealed it very tightly and a vacuum 
could be easily maintained for twenty hours. All dryings were made 
at 105 C and about 1.5 cm, pressure. At the end of the drying 



DAVTSSON & SIVASLIAN I MOISTURE IN SOILS. 



201 



period the vacuum was broken slowly and the ingoing air thoroughly 
scrubbed through concentrated sulfuric acid. The phosphorus 
pentoxide was placed in a shallow dish on a shelf about 3 inches 
below the shelf on which the samples were placed. 

The samples. 10 grams, were weighed into glass weighing dishes 5 
cm. in diameter and 3 cm. deep with ground-glass covers. Larger 
dishes, 7^ cm. X 3 cm. were compared with the smaller ones and 
found to be no better for a 10-gram sample. After drying, the dishes 
were closed and cooled in desiccators over calcium chloride and were 
weighed. It was found to be bad practice to place many samples in 
one desiccator because the opening of the desiccator to remove dishes 
permitted moisture to enter and those dishes remaining longest in the 
desiccator gave lower values for moisture. It is advisable to use an 
individual desiccator for each, sample when accurate work is desired. 
In no case should the bare hands be allowed to touch the dishes 
after being cleansed for the sample. 

Cornstarch was used for checking the methods because of its finely 
divided condition, high hygroscopicity, homogeneity, and its remain- 
ing unchanged during the drying process. The employment of such 
a substance makes it easy to establish if an absolutely dry substance 
can be obtained. The starch was uniformly mixed and preserved in 
glass-stoppered bottles. 

The data obtained for the moisture content of the starch sample 
by the vacuum method are reported in Table 1. The total moisture 
was removed in a period of two hours, when further drying did 
not yield any change in the moisture content. 



Table i. — Grams of moisture in 10 grams of starch, as shown by drying for 
different periods in a vacuum oven over P 2 5 at 103° C. 



Time of drying, 
hours. 


Number of determination. 


Average. 


1 


2 


3 


4 


5 


One 


I.27II 


1.2638 


I.265O 


1. 2361 


I.2603 


I.2592 


Two 


1.2897 


1.2938 


I.295O 


1.2960 


I.2950 


I.2939 


Three 


1-2953 


1.2962 


1-2954 


1.2902 


I.2889 


I.2932 


Four 


1.2948 


1. 2914 


1.2936 


1.2960 


I-29I5 


1-2934 



Ten determinations were then made to obtain the probable error 
of the method. The error calculated by the method of least squares 
is satisfactorily low and shows that very consistent results can be 
obtained. The determinations of moisture in grams in 10-gram 
samples of starch were as follows: 1.2963, 1.2947, 1.2942, 1.2953, 
1. 2941, 1.2926, 1.2927, 1.2936, 1.2888, and 1.2976, with an average of 



202 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



1.2939 and a probable error of ±0.0017, or a percentage error of 
±0.001. . 

A starch sample was then subjected to drying at the same tempera- 
ture but no phosphorus pentoxide was used to absorb the moisture. 
The vacuum pump was kept running all the time. On drying for 
two hours with this method a new sample of starch gave 1.2804 
grams of moisture. The same starch yielded 1.2203 grams of mois- 
ture in two hours and 1.2739 grams in 24.5 hours when no phos- 
phorus pentoxide was used. This shows that the moisture can not 
be removed easily without the use of the pentoxide and the time 
necessary for drying the sample is lessened greatly when the oxide is 
used. 

Drying in electric and gas ovens was then compared with the 
vacuum method. The data in Table 2 show the superiority of the 
vacuum method. The moisture obtained by direct drying is far 



Table 2. — Grams of moisture in 10 grams of starch as shown by drying in 
ovens of various kinds at a uniform temperature of 105 C. 



Kind of oven. 


Time of dry- 
ing, hours. 


Number of determinai ion. 


Average. 


X 


2 


3 


4 


5 


Vacuum, over P?Os 

Gas 


2.00 
16.50 
26.25 


1. 2817 
1. 2581 
1.22 I 3 


1.2766 

1-2534 
1. 2159 


I.2792 
I.2429 
1. 2210 


1. 28l6 
1.2556 
1. 2195 


I.2829 
1.2323 
1.2202 


I.2804 
I.2484 
I.2I95 



below that obtained by the vacuum method. The electric oven gave 
higher results than the gas oven and the results within themselves 
are not as concordant as those of the vacuum method. The superior- 
ity of the electric oven over the gas oven is shown and for this reason 
the gas oven was not employed in studying the moisture determina- 
tions on soils. 

The data obtained for comparison of the electric oven and the 
vacuum method as applied to soils are found in Table 3. Clay and 
muck -oils were used for this study because they present greater 
difficulties of moisture determinations than other soils, due to their 
large amount of surface and high hygroscopicity. The soils were 
air dried and ground to pass a 20-mesh sieve. Ten-gram samples 

yen used. The data show the superiority of this vacuum method. 
The electric oven gave results 2.2 percent less for muck and 3.4 per- 
cent less for clay than the total moisture found by the vacuum 
method. It is also shown that ,| hours' drying is sufficient to remove 
lire by the vacuum method, whereas the electric oven did 
not yield definite results in 10 hours' drying. 



DAVISSON & SIVASLIAN I MOISTURE IN SOILS. 



203 



Table 3. — Grams of moisture in 10-gram samples of muck and of clay soil, as 
determined by drying for different periods in a vacuum oven 
over P2O5 and in an electric oven. 

Muck Soil. 





Time of dry- 




Number of determination. 




















Kind of oven. 


ing, hours. 




? 


3 


4 


5 


Average 


Vacuum, over P2O5 


2.0 


1.8762 


1. 8814 


1.8892 


I.8806 


I.8792 


I.8813 


Do. 


4-0 


1-9035 


1. 9125 


1.9016 


I.9032 


I.8994 


I.9040 


Do. 


7.0 


1-9055 


1. 9148 


1.9088 


I.9087 


1.9022 


I.9080 


Electric 


4.0 


1.7764 


I.8403 


1. 8491 


1.8332 


I.8296 


I.8257 


Do 


7.0 


1.8200 


I-8563 


1. 8612 


1. 8612 


1. 8619 


1.8538 


Do 


IO. 


1. 8210 


1-8533 


1-8739 


1.8732 


I.8856 


1. 8612 


Clay Soil. 


Vacuum, over P2O5 


3-0 


o.445i 


0.4451 


0.4412 


O.4422 


0-4435 


0-4434 


Do. 


5-o 


.4471 


•4473 


•4443 


•4525 


•4445 


.4471 


Do. 


7-5 


.4467 


•4473 


•4434 


•4477 


•4425 


•4455 


Electric 


5-0 


.4191 


•4135 


•4293 


.4227 


.4150 


.4199 


Do 


8-5 


.4211 


•4305 


•4341 


•4234 


.4270 


.4272 


Do 


11. 


.4215 


•4355 


.4306 


•4309 


.4278 


.4292 



The vacuum method produced a dry soil, a condition which can 
not be obtained by the electric oven on several hours' drying. The 
danger of oxidation of the organic matter is removed when the 
vacuum method is employed. The loss of any volatile matter from 
drying in the vacuum oven will probably not be any greater than the 
loss which will take place on long drying in the gas or electric oven. 

The air-drying oven is not capable of removing all the hygroscopic 
moisture from a soil sample. Soils having a high hygroscopicity will, 
therefore, retain considerable moisture under the ordinary procedure 
of drying and the results obtained will not represent the actual mois- 
ture content of the sample. 

In order to show that the electric drying oven does not remove all 
the moisture from a soil sample, a determination was made by taking 
samples of muck soil and drying for 3^2 hours in the vacuum oven 
over P 2 5 at 105 C. The dry substance found was 8.0750 and 
8.0978 grams for the two samples. These samples were then placed 
uncovered in a desiccator over 10 percent sulfuric acid and the 
desiccator evacuated. After standing in this desiccator for 15 hours 
the samples w r ere removed and dried in the electric oven at 105 C. 
for 7.5 hours and the dry substance found was 8.1800 and 8.1688 
grams. The samples were again placed over 10 percent sulfuric 
acid in a vacuum and after 15 hours they were dried by the vacuum 
method for 3.5 hours at 105 C, when the dry substance found was 



204 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



8.0795 and 8.0938 grams, respectively. The values obtained by the 
electric oven are considerably higher than those obtained by the first 
drying of the soil sample by the vacuum method. The oven is, 
therefore, not capable of removing all the hygroscopic moisture from 
the soil. The second drying by the vacuum method gave results 
which are very nearly the same as those obtained by the first drying. 
This method of drying gives results which are definite and represent 
the actual moisture content of the soil. - Soils having high hygro- 
scopicity offer no difficulty in moisture determinations by the vacuum 
method, while direct drying fails to remove all the hygroscopic mois- 
ture. 

Conclusions. 

1. Drying of soil samples in gas or electrically heated ovens will 
not give the true moisture content of the soils. 

2. Drying by the vacuum method yields trustworthy and con- 
cordant results. 

3. Four hours' drying at 105 C. in a vacuum over phosphorus 
pentoxide is sufficient to remove the moisture from soils having high 
hygroscopicity. 




BOSHNAKIAN : SHAPE OF THE WHEAT KERNEL. 



205 



THE MECHANICAL FACTORS DETERMINING THE SHAPE 
OF THE WHEAT KERNEL. 1 

Sarkis Boshnakian. 

The shape of the grain of wheat is affected by a number of 
spikelet characters, which are mainly : (1) The stiffness of the glumes, 

(2) the size and shape of the space in which the grain develops, 

(3) the number of grains in the spikelet and their position, (4) the 
density of the head, (5) the pressure caused by the growth of dif- 
ferent parts of the head, and (6) the species which produces the 
kernel. Just as the interior surface of stone or hard-shelled fruits 
determines to a very large extent the shape of the enclosed seed, so 
the character and form of the surroundings of the developing grain 
of wheat determine the shape of the mature kernel. 

An ideal wheat kernel whose free development has not been 
arrested by coming in contact with the surrounding parts of the 
spikelet should be symmetrical ; that is, when the grain is divided 
by a plane dorsi-ventrally passing through the crease, the two halves 
should be alike with the dimensions reversed. Such a grain is shown 
at A in figure 27. The position in the spikelet of uniformly de- 
veloped kernels is seen in figure 27 E. Symmetrical grains are very 
rarely found in nature. 

The types of kernel most frequently found are those shown in B 
and C of figure 27 ; their relative positions on the spikelet are shown 
in Fb and Fc respectively. In these cases the plane does not divide 
the grain into symmetrical halves. Grain B dorsally viewed is 
flattened on the left side, while in the case of C the opposite side is 
flattened. This flattening occurs always on the side of the kernel 
which is nearer to the rachis, as shown in F. It is evident then 
that all grains which are flattened on the left side are produced by 
the florets on the left side of the spikelets, and those flattened on 
the right side are borne on the opposite side of the spikelet. This 
character, which is constant for all species, enables one to determine 
the side of the spikelet on which different kernels were produced. 

1 Received for publication November 6, 1917. 



206 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



All species of wheat do not produce this type of flattening in the 
same degree. In Triticum vulgare, although easily noticed, it is not 
so pronounced. The writer's observations of the various forms show 
that they rank about as follows with respect to their degree of 



flattening 



Triticum vulgare 
T. capita-turn; 
T. compaction; 
T. polonicum ; 
T. turgidum; 



6. Triticum durum; 

7. T. spelt a; 

8. T. dicoccoides; 

9. T. die oc cum; 
10. T. monococcum. 



In durums the degree of flattening varies with different varieties. 

The extent to which flattening takes place depends primarily upon 
the stiffness and shape of the outer glumes ; the stiffer the glumes 





4fe 
H 



V 




Fig. 27. — Types of wheat kernels. Aa, symmetrical grain ; Ab, cross section 
(m-m\ axis). Ba, asymmetrical grain produced on left side of spikelet, left 
side of kernel flattened ; Bb, cross section. Ca, asymmetrical grain borne on 
right side of spikelet, flattened on right side ; Cb, cross section. E, cross 
section of spikelet showing position of well-developed symmetrical kernels; 
a, grain; b. palca ; c, fertile glume; d, outer glume; c, rachis. F, position of 
grains with rounded cheeks, glumes not shown ; note flattening of side of 
rachis, c. (i, grains growing close together, showing flat cheeks. //, spikelet 
with the kernels tightly held together, showing effect of central spikelet b on 
the lhape of tin cheek* Of the lateral kernels a and r; flattening of checks, 
as in <i, taking place when grains b and c face ventrally, depression of crease 
taking plaOC when the lateral kernel a faces the dorsal portion of the central 
kernel b. I, Iterilc or partially developed central floret producing slight de- 
pressions along the crease of the lateral florets b and c. J, spikelet of T. 
mo tun fa cum showing grain with protruded cheeks; r, rachis; /, sterile floret. 
A' and /.. effed of position of central grain b on the depression of crease of 
lateral grain a. K, depression on lower part of grain a. L, depression on 
upper portion of grain a. 



BOSHNAKIAN : SHAPE OF THE WHEAT KERNEL. 



207 



the more tightly the grains are pressed toward the rachis. On the 
side of the rachis the grain does not have sufficient room for de- 
velopment and therefore does not fill up as fully on this side as it 
does on the other. The ranking shown above is in reality a ranking 
of the species according to the stiffness of their outer glumes. 

The cheeks of the grain are the visible parts on each side of the 
crease when the kernel is viewed ventrally. The cheeks may be 
round and plump (F), flat (G), sunken with sharp edges (I), or 
protruded (/). The cheeks are filled, plump, and round (F and F) 
when the spikelet is soft and spreading, and contains not more than 
three grains. This type of a roomy spikelet will be found among 
the vulgare and squareheads, and in some of the clubs. The flatten- 
ing of the cheeks (G) takes place when there are two kernels which 
face each other in the spikelet and when the latter is stiff and tends 
to press these two kernels together. The flattening of the cheeks in 
this case is mechanical. 

The reader should be reminded that the grain of wheat throughout 
its period of development is very soft and that it hardens only after 
it attains its maximum development. Hardening is a drying process 
and occurs during the last few days of its period of maturation. As 
the kernel is very soft before this period, the slightest pressure on 
the grain through contact is very apt to modify its shape. 

The process of the flattening of the cheeks of the wheat grain 
is not different from the flattening of the sides of the horse-chestnut. 
It will be remembered that when only one chestnut is developed from 
a flower it has more or less rounded sides ; when, however, the flower 
produces two chestnuts the sides which face each other are flattened 
out instead of being rounded as in the first instance. We find the 
same phenomenon in maize. In well-fertilized ears the four sides 
of the kernels are flat, while in poorly fertilized ears where an isolated 
pair of kernels are found the sides of the grains which face are flat 
while the sides which do not come into contact remain rounded. This 
principle of flattening in this and in many other cases is due to 
pressure resulting through contact. Flat cheeks will be found as a 
rule among the 2-grained and tight-glumed species such as the wild 
wheat, the emmers, and most of the spelts. 

When three grains are present (/) and the texture and form of 
the glumes are as in the previous case the cheeks of some of the 
kernels of the basal florets tend to sink or flatten, depending on the 
size and position of the seeds developing between the two basal 
grains. If the central grain is small (lb), the cheeks of the large 



2o8 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



kernels on both sides (7a and Ic) of this tend to sink to fit the out- 
line of the central grain (lb). If on the other hand the middle grain 
is fairly large (Hb), usually the grains whose ventral sides face one 
another (Hb and He) are flattened ventrally more or less while the 
other outer grain (Ha), which faces the dorsal portion of the cen- 
tral kernel {Hb), develops a depression along the crease and the 
edges of the cheek become sharp. The presence of sterile florets in 
the center will produce a slight depression along the ventral surface. 

The position of the central grain or grains will determine the region 
where the depression is to occur. If the small central grain (Kb) 
is located along the middle portion of the crease of the outer kernel 
(Ka) the depression occurs along that region. If it has a higher 
position (Lb) the upper portion of the cheek only will be depressed 
(La). Grains of this third type occur particularly among spelts and 
often in durums and English wheats. 

The fourth and last case is that of the monococcum, whose spikelets 
almost always produce but one kernel. If the glumes were not stiff 
and did not press the grain inside tightly against the rachis the shape 
of the normally developed grain might approach that of the vulgar e. 
On account of the pressure along the sides of the grain the latter 
naturally flattens, and since there is no other grain in the same 
spikelet which by contact will arrest the growth of the cheeks the 
latter keep on protruding until the grain attains its full size (/). 

That considerable pressure is exerted by the glumes may be demon- 
strated by the presence of parallel vein marks on the sides of the 
grain of monococcum, especially that facing the rachis. The parallel 
sunken lines are the exact reproductions of the veins of the lemma 
that envelops the grain. The vein marks of the glumes could not 
have been impressed on the plastic grain without pressure and con- 
tact. Such vein impressions are also occasionally seen on the kernels 
of the spelts. 

The shape of the kernel of the club wheats is often different from 
that of the Others. The grains of this form are short and in some 
cases .almost spheroid, depending of course upon the variety. The 
ihortnen of the grain is caused by the presence of a genetic factor or 
:..< ton which shortens a number of size characters such as length 
of glumes, awn*, intcrnodes, culm, and kernel. This is not a phys- 
ical factor. 

In club wheats we find, besides shortness, an irregularity as re- 
gards tin: liapc of the grain. Kernels obtained from the central 
portion of tin- bead of most of the clubs are fairly symmetrical and 



BOSHNAKIAN : SHAPE OF THE WHEAT KERNEL. 



209 



uniform, but those taken from the upper portion of the head, espe- 
cially of squareheaded clubs, are irregular and often flattened some- 
what like those of Triticum monococcum. This is due to the close 
and dense arrangement of the spikelets one above the other. When 
the internodes are short the spikelets spread out and arrange them- 
selves almost at right angles with the rachis one above the other. 
Under such conditions the development of a kernel is checked 
through the growth of competing grains in spikelets above and below 
it. The pressure developed from growth acts along a direction 
perpendicular to the plane of symmetry of the kernel and tends to 
flatten out the cheeks. The irregularities are produced by pressure 
as well as by the crowding of the grain on the sides, above, and 
below each developing grain. 

There are other mechanical causes affecting the shape of the grain ; 
though mechanical they are not due to pressure. Irregularities 
on the surface of the grain are to a great extent the result of evapora- 
tion of the moisture in the grain. We have soft and hard grains 
with numerous intermediate gradations. Soft or starchy grains have 
almost always rounded cheeks ; whereas those which are hard or 
corneous shrivel to some extent upon drying and consequently pro- 
duce irregularities on the surface of the grain. This phenomenon is 
comparable with the smoothness of the surface of the soft starchy 
and the wrinkled surface of the sweet or dent corn. 

The mechanical causes determining the shape of the grain should 
not be confused with the purely genetic factors which are to some 
extent responsible for the production of certain grain forms. Refer- 
ence has already been made to the genetic causes affecting size of 
the kernels of dense wheats. The shortness of the grain of the 
club and the unusual length of that of the Polish wheat are not the 
results of any form of mechanical pressure but are already deter- 
mined before the embryo begins to develop. The grain will remain 
short or grow long depending upon the species and the length factor 
which the plant carries. 

There are many grain characters which are the result of both 
genetic and mechanical factors. It was said at the beginning of the 
discussion that the shape of the grain was primarily dependent upon 
the shape and stiffness of the glumes'. The shape of the glumes 
varies with different species, and the grain whose shape is to a 
great extent dependent on that of the glumes varies also depending 
upon the species. The first cause in this case is genetic, but its 
result on the grain is mechanical. Most of the forms considered in 
this paper belong to the latter category. 



210 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



LOSS OF ORGANIC MATTER IN CLOVER RETURNED TO 

THE SOIL. 1 

George E. Boltz. 2 
Introduction. 

The practice of applying organic matter is generally considered 
more or less essential in maintaining and increasing the fertility of 
certain soils. The value of such soil treatment is supposed to be 
derived largely from the beneficial effects of the organic matter as 
well as the fertilizing elements incidentally carried by it. In many 
instances the fertilizing elements are considered only of secondary 
importance, while it is assumed that the organic matter in crops and 
manures is the chief factor which causes the increased productive- 
ness of soils. 

It is often advocated in agricultural literature that the most effi- 
cient method of maintaining or increasing the supply of organic 
matter in the soil is by the use of green manures and crop residues. 
This method is given preference over that of feeding the crop and 
returning the manure to the soil, because a large percentage of the 
organic matter is destroyed in passing through the animal. With 
clover this loss represents about 66 percent of the total organic matter 
present in the crop. By incorporating the green crop directly in the 
soil or by cutting and allowing it to remain on the surface for some 
time before plowing under, it is assumed that this loss is greatl> 
reduced if not entirely avoided. 

In order to obtain information regarding the loss of organic matter 
from a clover crop used for green manuring as compared with the 
]<>- when fed to farm animals and returned to the soil in the 
form of manure, experiments were conducted at the Ohio Agricul- 
tural Kxpcrimrnt Stat inn w ith the results here given. 

1 Contribution from tlx- Ohio Agricultural Kxperiment Station, Wooster, 
Ohio. Publication Approved by the Director. Received for publication Novem- 
ber IJ, 1017. 

1 The author wishes to acknowledge the many useful suggestions made and 

'i ' :<!.''«,, ,| rluriiH' the rout <<• of this investigation by C. G. Williams, 

Igrooomiat, ftnd J. W. Ames and C J. Schollcnbcrgcr, chemist and assistant 
'ii'iiiist, all of the Ohio station. 



BOLTZI LOSS OF ORGANIC MATTER IN CLOVER. 211 

Plan of Experiment 1915-1916. 

A plot of soil having an area of 18 square feet was treated with 
green clover at the rate of 7,744 pounds per acre, to correspond to 
the practice of mowing a crop and allowing it to remain on the sur- 
face of the ground for some time before plowing it under. A similar 
area was treated with a like quantity of clover, which was spaded 
under. A similar experiment was conducted in a lysimeter test, each 
container having an area of 9 square feet and receiving an applica- 
tion of green clover at the rate of 17,520 pounds per acre. The loss 
on ignition of the moisture-free soil and clover was determined on 
samples taken at the beginning and end of the experiment, and the 
loss of organic matter calculated from these data. The clover was 
applied to the soil October 12, 191 5, and the residue collected and 
soil sampled on May 5, 191 6, so that the clover was exposed to the 



Table i. — Organic matter retained in soil at the end of 206 days and percentage 
lost from clover left on surface as compared with that spaded 
into the soil, 1915-16. 



Method of application 


Clover 
applied. 


Residue on 
surface at 

end of 
experiment. 


Applied. 


Organic matter. 

In residue . 
at end of Retained 
experiment. b y soll> 


Loss. 




Pounds. 


Pounds. 


Pounds. 


Pounds. 


Pounds. 


Percent. 


Clover on surface. 














Field test 


2,000 


6392 


1,075.4 


360.9 


4.1 


66.05 


Lysimeter test 


2,000 


440.1 


1,202.8 


302.5 


76.0 


68.52 


Clover mixed with soil. 














Field test 


2,000 




1,075.4 




769.4 


28.45 


Lysimeter test 


2,000 




1,202.8 




498.8 


58.53 



action of the weather for a period of 206 days. The figures in Table 
1 express the loss of organic matter during the experiment of 191 5~ 
16 in a ton of green clover when applied at the rate indicated above. 

Plan of Experiment 1916-1917. 

Two sets of duplicate plots, adjacent to each other and each 
measuring 6 by 3 feet, were spaded to a depth of 6 inches ; during 
the spading, 1,573 grams (3.46 pounds avoirdupois) of dried and 
finely cut clover were thoroughly and evenly mixed with the soil of 
each of two plots, while the same quantity of uncut clover was spread 
upon the surface of each of two others, the soil having been com- 
pacted after spading. This quantity on an area of 18 square feet 
corresponds to 8,000 pounds or 4 tons of dry clover to the acre. 
Samples of both the clover used and of the soil from each plot before 



212 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY 



adding the clover were taken and subsequently analyzed. The plots 
were prepared as described on November 8, 191 6, and were imme- 
diately covered with a wire screen in order to prevent the clover 
spread on the surface from being blown away, as well as to prevent 
other material from being carried on to the plots. The duration of 
the experiment was 187 days, from November 8, 191 6, to May 4, 
1 91 7, when the plots were sampled in the same manner as at first, 
after carefully removing the residue of clover from the surface of 
the two plots on which it had been spread. 

The first experiment was intended to furnish data on the loss of 
organic matter only ; in order to make the results of the second ex- 
periment of greater value, carbon and nitrogen determinations on the 
different samples were made instead of the determination of loss on 
ignition. 

Analyses were made of the soil from each plot, both at the begin- 
ning and at the end of the experiment. The clover applied and the 
residue removed were also analyzed. The percentages of carbon and 
of nitrogen in the soil at the beginning and at the end of the experi- 
ment are shown in Table 2, while Table 3 gives data on the carbon 



Table 2. — Percentage of carbon and of nitrogen in soil to which 8,000 pounds 
of dry clover to the acre was added on the surface and by spading 
in. at the beginning and at the end of a period of 187 days. 



Element and time of 
determination. 


Clover spaded in. 


Clover spread on surface. 


Plot 1. 


Plot 3. 


Average. 


Plot 2. 


Plot 4. 


Average, 


Carbon: 














At beginning 


I.O58 


1. 021 


I.O39 


1. 021 


I.087 


I.054 


At end 


1. 148 


1. 1 69 


1. 158 


I.064 


1. 041 


I.052 


Nitrogen: 














At beginning 


.097 


.097 


.097 


.097 


.097 


.097 




.105 


.105 


.IO.S 


.099 


.!()() 


.099 



and nitrogen in pounds per acre. The clover added contained 42.31 
percent of carbon and [.99 percent of nitrogen. From Plot 2 6,164 
Ul dl of residue wen- removed, containing 29.34 percent of carbon 
and [.91 percent of nitrogen, while from the corresponding Plot 4 
7,568 pounds of residue were removed, containing 23 percent of 
carbon and 1 .3X percent of nitrogen.. 

\<] M I.IS tth Kxi'KKIMKNTS. 

There Mrai Vtty little if any losa of nitrogen during the experi- 
ment, either from the clover incorporated or left on top of the soil. 

< taring to the ffiiall increase in percentage of total nitrogen produced 



BOLTZ : LOSS OF ORGANIC MATTER IX CLOVER. 



213 



in the soil by the clover added and to the difficulty of securing a rep- 
resentative sample as well as the limitations of the method tor total 
nitrogen in soils, the slight differences shown in the table of the 
nitrogen recovered are well within the limits of analytical error. 

Table 3. — Pounds per acre of carbon and of nitrogen in soil to which 8,000 
pounds of dry clover to the acre was added on the surface and by 
spading in, with loss from the clover at the end of 187 days. 

Carbon. 



Clover spaded in. Clover spread on surface. 



.Determination. 


Plot I. 


Plot 3. 


Average. 


Plot 2. 


Plot 4. 


Average. 


At beginning 


19.785 


19.093 


19.439 


19.093 


20,327 


19.710 


Added in clover 


3.385 


3.385 


3.385 


3.385 


3.385 


3.385 


Total 


23.170 


22,478 


22,824 


22,478 


23.712 


23.095 




21,468 


21,860 


21,664 


19,897 


19,467 


19,682 


Loss during period .... 


1,702 


618 


I,l6o 


2,581 


4.245 


3.413 


Remaining in residue . . 








1,809 


1. 741 


1,775 


Net loss 


1,702 


618 


I,l60 


772 


2.504 


I.638 


XlTROGEX. 


At beginning 


1,814 


1,814 


1,814 


1,814 


1,814 


1,814 


Added in clover 


159 


159 


159 


159 


159 


159 


Total 


1.973 


1-973 


1.973 


1.973 


1.973 


1.973 




1,964 


1,964 


1,964 


1,851 


1,870 


i,86o 


Loss during period .... 


9 


9 


9 


122 


103 


113 


Remaining in residue . . 








118 


104 


112 


Net loss 


9 


9 


9 


A 


°i 


1 



a Apparent gain. 



However, the indications are that no appreciable loss of nitrogen oc- 
curred. 

The data in Table 1 show that the organic matter in a ton of 
green clover when cut and allowed to remain on the surface of the 
soil for 206 days decreased from 1,075 pounds originally present to 
361 pounds. Of the 715 pounds of organic matter that disappeared 
from the surface a mere trace, 4 pounds, was retained by the soil 
while 711 pounds were lost through processes of decay. The rate of 
application and the stage of maturity of the clover used in the lysim- 
eter test were different from those used in the field test, but the per- 
centage loss is practically the same in both cases, being 66.05 percent 
for the field test and 68.52 for the lysimeter test. 

The loss of carbonaceous matter expressed as elementary carbon 
was greater when the clover was allowed to lie upon the surface than 
when turned under in both the 1915-1916 and 1916-1917 experi- 
ments. The average loss shown by the two plots of 1916-1917 with 



214 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



the clover on the surface was 48.38 percent, while those with the 
clover incorporated in the soil lost 34.26 percent of the carbon ap- 
plied during the same experiment. The results of the 1916-1917 
experiment are similar to those obtained in 1915-1916. Differences 
in season and duration of experiment would undoubtedly affect the 
results. 

Value of Green Manures. 

Green-manure crops assimilate, concentrate, and return to the soil 
fesifiajjal plant nutrition elements that are available for succeeding 
cv^^T The fertilizing elements thus prepared for assimilation by 
other crops may have a beneficial influence after the greater por- 
tion of the organic matter by which they are carried is destroyed. 
Half or more of the organic matter in a crop, depending upon the 
method employed in returning it to the soil, may be destroyed within 
six months, yet the beneficial effects are apparent for several years 
afterward. That mineral fertilizers have a decided value in crop 
production, whether accompanied by organic matter or not, can not 
be disputed. Consequently, the value of the fertility elements in a 
green-manure crop should not be ignored, although they originally 
came from the soil. Crops having high fertility value should be re- 
turned to the soil for their fertilizing constituents as well as for any 
theoretical value of the organic matter. 

Advantage of Feeding Crop and .Applying Manure. 

Data reported in Ohio station Bulletin 183 from an experiment in 
which 28 steers were fed on a cement floor, the duration and season 
of the experiment corresponding very closely to the plot experiments 
under discussion, indicate that almost exactly one third of the total 
organic matter in the feed and straw used for bedding was recov- 
ered in the manure. This experiment shows that practically as much 
Organic matter is applied to the soil when a crop of clover is fed and 
the manure applied to the soil as when the crop is allowed to remain 
on the surface of the ground from fall until spring before being 
plowed under. 

Eliminating the comparatively small amount of fertilizing elements 
lot( in metabolic processes when feeding clover to farm animals and 
Considering the organic matter only, very little would appear to be 
gained by plowing a Clover crop under rather than feeding it and ap- 
plying the manure. 



LE CLERC, BAILEY AND WESSLING '. BAKING TESTS. 



215 



MILLING AND BAKING TESTS OF EINKORN, EMMER, 
SPELT, AND POLISH WHEAT. 1 

J. A. Le Clerc, L. H. Bailey/ and H. L. Wessling. 

Wheats are classified according to Hackel as follows : 

Triticum monococcum — Einkorn or one-kerneled wheat. 
Triticum sativum die oc cum — Emmer. 
Triticum sativum spelta — Spelt. 
Triticum sativum tenax vulgare — Common wheat. 
Triticum sativum tenax durum — Durum wheat. 
Triticum polonicum — Polish wheat. 

The einkorn used in our investigations (C. I. No. 2433) was a 
small grain irregular in size which did not separate readily from the 
chaff. It seemed much like an immature and shriveled sample of 
ordinary wheat. Some of the grains were so small that they con- 
tained very little endosperm. 

The kernels of the emmer (Black Winter, C. I. No. 2337) which 
was used in this experiment were larger than ordinary wheat and in 
shape resembled somewhat the rye grain. It was also difficult to 
remove the chaff from the grain. 

The sample of spelt 2 (Alstroum, C. I. No. 3264) resembled emmer 
quite closely in appearance of the grain ; in fact, the names are some- 
times confused in this country. Spelt and emmer are both used 
mostly as stock feed and very little for human food. 

The sample of Polish wheat used was obtained from C. B. West, 
Sheridan, Wyo. The kernels of this sample were even larger than 
those of durum wheat and about twice as long as those of ordinary 
wheat. In appearance the kernels were flinty and of an amber color, 
thus resembling durum wheat. Polish wheat differs from the ein- 
korn and spelt in that it is readily separated from the chaff, so that 
ordinary thrashing is all that is required to prepare the grain for the 
miller. 

1 Contribution from the Laboratory of Plant Chemistry of the Bureau of 
Chemistry, U. S. Department of Agriculture, Washington, D. C. Received for 
publication April 3, 1918. 

2 The three wheats (einkorn, emmer, and spelt) were obtained for us through 
the courtesy of the Office of Cereal Investigations, Bureau of Plant Industry. 
The emmer and spelt were supplied hull free. The sample of einkorn was 
hulled in the laboratory before milling. 



2l6 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Of the wheats, Triticiim sativum tenax vulgar e is the usual source 
of flour for bread and biscuits. Durum wheat is especially adapted 
for the manufacture of semolina used in making macaroni. The 
other wheats have seldom been used in the manufacture of flour. 
Emmer is sometimes used as a breakfast food. 

The object of this experiment was to determine whether these 
rarer wheats could be successfully milled into good flour and whether 
the flours produced therefrom would make good bread. Samples of 
all of these wheats were milled on an Allis-Chalmers experimental 
flour mill, a sample of hard spring wheat being likewise milled at the 
same time as a check. All the samples were tempered to 13 percent 
of moisture over night, then scoured and tempered to 15 percent 
for 2 hours, and then milled so as to obtain a 65 percent patent flour. 
Xo attempt was made to obtain a straight grade of flour or to obtain 
a maximum flour yield. 

The results of milling showed that Polish wheat and emmer have 
a smaller amount of bran than hard spring wheats, thus indicating 
a possibly larger flour yield. Even when milled on a 65 percent 
patent basis all of these flours (except that from spring wheat) were 
somewhat gray in color, although the flour from spelt was only 
slightly darker than that from spring, wheat. 

Table 1 shows the composition of the flours and gives the charac- 
teristics of the breads produced therefrom. 

From this table it will be seen that flour made from spelt is more 
nearly like that from spring wheat in its ash, nitrogen, acidity and 



Tabu: i. — Composition of flours and characteristics of bread made from 
einkorn, emmer, spelt, and Polish wheat. 

COMPOSITION OF FLOURS. 



Hard spring 
Character. wh £ u 


Einkorn. 


Emmer. 


Spelt. 


Polish 
wheat. 


Maximum rxpntision of 'lough, c.c. 


II.750 
393 

10.090 
0.075 

27. [00 

0.000 
69.500 
780 


10.450 
0.811 
15.500 
0.223 
37.700 
14.000 
66.500 
320 


I 1 . 200 

783 
13.790 

O.213 
44.800 
1 4.6OO 
82.OOO 

680 


I I. OIO 

0-555 
12.250 

O. 151 
31.700 
I I. OOO 
67.OOO 
680 


10. v>o 
O.848 

1 3- 050 
0.325 
34- 500 
1 2.600 
76.000 
640 


CHARACTERISTICS OF BREADS. 




07.5 c. 1 


v. blown 
96 

<)2 


94 <--K- 2 

95 

08 


96 e.g. 

96 

98- 5 


95 R- 
04 

99 



c — creamy. k — Rray. 



LECLERC, BAILEY & WESSLING : MILLING TESTS. 



217 



gluten content than the other flours. Flour from einkorn, emmer, 
and Polish wheat are very high in ash, acidity, and in gluten. 
Emmer flour showed an extremely high absorption capacity, while 
Polish-wheat flour was second in this respect. Flours from spelt and 
einkorn had approximately the same absorption as that possessed by 
spring-wheat flour. 

When these flours were baked into bread, the bread made from 
the einkorn was the least desirable not only in volume but in color, 
elasticity, and in taste or flavor. In fact, the results obtained with 
einkorn might indicate that this kind of wheat is not suited to make a 
good yeast-risen bread. That made from emmer was somewhat 
better, although not equal to bread made from spring-wheat flour in 
color, volume, and elasticity. The Polish-wheat bread had a better 
color than the emmer bread, but the volume was not quite so large. 
The crumb, however, seemed more moist than that from other breads 
and in fact this bread had an elasticity almost equal to that of spring- 
wheat bread. The spelt bread on the other hand compared favorably 
in most respects to the ordinary wheat bread, except that the color 
was a little darker. There was no difference in taste and appearance 
from ordinary wheat bread. 

Summary. 

1. The rarer wheats, emmer and spelt (both free of hulls) and 
Polish wheat, can be milled into a satisfactory flour and the flour 
used in baking a good loaf of bread. 

2. The results obtained from einkorn (free of hulls) are not so 
encouraging. 

3. Alstroum spelt seems particularly adapted to the production of 
a good flour for baking purposes. 

4. Black Winter emmer flour has a very high absorption capacity. 

5. In view of the well-known present deficiency in our wheat 
supply . the use of emmer, spelt, and Polish wheat as human food 
(bread, breakfast cereals, etc.) should be encouraged wherever they 
are available. As they are not superior to the more common varie- 
ties of wheat there would be no advantage in having them replace 
ordinary wheats in normal times. 



2i8 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



COMPARATIVE SMUT RESISTANCE OF WASHINGTON 

WHEATS. 1 

E. F. Gaines. 

Damage Caused by Stinking Smut in the Northwest. 

Washington, Oregon, and Idaho produce over 40,000,000 bushels 
of winter wheat annually. Stinking smut is more prevalent in the 
winter-wheat sections of these States than anywhere else in the 
United States, and probably a larger percentage of the crop is affected 
than on an equal acreage anywhere else in the world. It is not un- 
common to find whole fields with 40 percent of smut. Several county 
I s in the more important winter- wheat growing counties have 
estimated that these three States lose in the neighborhood of 15 
percent of their winter wheat by smut. This would mean a loss of 
6,000,000 bushels, which are worth, at present prices, over $10,- 
000,000. The prevalence of smut in winter wheat seems to be in- 
creasing in spite of the most careful methods of seed treatment. 
This condition is causing many farmers to abandon winter wheat 
until some measures can be found to control smut. The elimination 
of winter wheat is undesirable from the standpoint of distribution 
of labor in seeding and harvesting and from the fact that winter 
wheat produces from 1 to 5 bushels per acre more than spring wheat. 

['resent Methods of Smut Control. 

Several methods of producing winter wheat with little or no smut 
are known. It has been found both by experiment and by farm 
practice that little or no snint is produced when the crop is seeded 
very early in the season, but this is not desirable on account of 
overdi velopmenl in the fall and a consequent reduction in yield. 
Moreover, in .1 dry season or when the cultivation has not been the 
be '. the moisture in the soil in July or early August is insufficient to 
germinate the seed. It is impossible to sow early except on summer 
fallow. There is also the added expense of carrying over the seed 
from the previous year's crop. 

Little or no Hunt produced in fields ihat are so\Vn abnormally 

» Conti il.uf inn from tin- Washington Agricultural Kxpcrimcnt Station, Pull- 
man, Wash. Received for publication April 1, i«;iH. 



GAINES : SMUT RESISTANCE OF WHEAT. 



119 



late, but the weather between Thanksgiving and Christmas is so 
unsettled that it is often impossible to seed during this period. There 
is also danger of the grain freezing out when it is sown so late. 
Therefore, both very early and very late seeding, although eliminat- 
ing smut to a large degree, may be dismissed as undesirable as a 
general practice. Winter wheat is normally seeded during the 
months of September and October in the Northwest. This has been 
shown by repeated experiments to be the period of maximum infec- 
tion. It has been suggested that by replowing the summer fallow 
just prior to seeding, the wind-borne smut spores would be buried 
beyond the infection zone of the germinating grain and a reduction 
in smut would result .even when the sowing was done in the normal 
season. This practice is untenable because the expense is prohibitive 
and the time is not available at this season of the year. Moreover, 
it leaves the soil in bad physical condition. 

As none of the methods of smut control mentioned above are 
generally practicable, it was thought that, if a variety of wheat could 
be found that was immune or highly resistant to the attacks of this 
fungus, it would help in the elimination of this evil. 

Investigation ox Resistance: Method and Material. 

For the past four years the Washington station has been studying 
the comparative smut resistance of the most common winter wheats 
of the Northwest. The method employed was to blacken the seed 
with smut spores just before planting, and sow all the varieties on 
the same date, under uniform conditions, in rows 18 inches apart. A 
furrow was made with a small garden push plow, the seed spaced 5 to 
7 inches apart therein, and covered with the same implement by 
plowing in the earth from each side. Enough cultivation was given 
to keep down weeds in spring and early summer. At harvest time 
each variety was pulled as it became ripe. The plants were divided 
into three piles as follows : Plants smut free, plants all smutted, and 
plants partly smutted. The number of plants in each pile was re- 
corded, then the plants partly smutted were further divided into 
healthy heads and smutted heads and the number of each recorded. 
When a head was found that was only partially smutted, if it was 
more than half smutted it was put into the smutted pile ; otherwise it 
went into the uninfected pile. Table 1 gives the sum of these counts 
during the past three years. 

Table 1 gives the total infection distribution in tabular form of 
thirteen varieties during 191 5, 1916 and 1917. The plantings were 



220 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Table i. — X umber of plants smut free, all smutted, and partly smutted under 
conditions of maximum infection in 1915, 1916, and 1917, with the number 
of uninfected and smutted heads and the average number of heads 
produced by the partly smutted plants. 



Variety. 


Number of plants. 


Number of heads on partly smutted 
plants. 


Smut 
free. 


All smutted. 


Partly 
smutted. 


Not 
smutted. 


Smutted. 


Average 
total number 
per plant 




8lO 




298 


7,518 


5H 


26.9 




35 




8 


IOI 


17 


I4.7 




118 


7? 


305 


1,732 


4,45 2 


20 3 




85 


115 


320 


1,863 


5,754 


23-8 




10 


15 


122 


1,069 


2,744 


313 




77 


266 


539 


2,59i 


10,690 


24.6 




25 


149 


196 


949 


3,217 


21.3 


Little Club 


50 


304 


54 


233 


997 


22.8 




49 


575 


217 


910 


2,842 


17.3 




43 


496 


127 


606 


2,323 


23.I 




17 


244 


214 


630 


4,081 


19. 1 


Hybrid 108 


41 


497 


105 


490 


i,798 


21.8 


Hybrid 128 


43 


696 


139 


565 


2,467 


21.8 



made on November 7, October 14, and November 24, respectively. 
W hile this is later than winter wheat is usually seeded, yet the re- 
sults should be comparable, for all varieties received exactly the 
same treatment. The large number of heads per plant is not un- 
usual considering the cultivation they received and the fact that each 
plant had a space of approximately 108 square inches. Counts from 
an average of 575 plants of each variety or a total of 7,481 plants, of 
which 2,644 partly smutted plants were divided into 19,257 wheat 
heads and 41,896 smut heads, are given in Table 1 as the basis of the 
percentage comparison presented in Table 2. 

This compilation has been made in order to show in easily com- 
parable form, the susceptibility to smut of the different varieties of 
winter wheat. Hie first column, percentage of infected plants, is the 
sum of the plants all smutted and partly smutted in terms of per- 
centage of the total number of plants. The third column, total 
percentage of smut produced, is the sum of the smut produced on 
both tlx- plants partly smutted and wholly smutted. The fourth 
eohtmtl i- tin- percentage of smut produced in an entirely different 
fxjrf-rimcnt. It is based on a head count of part of one drill row in 
eadl plot of the field variety tests and is the average percentage of 
Ifnutted heads produced during the last three years from treated 

grain. I his column represents the percentage of smut produced in 
winter wheal under conditions as unfavorable for smut production 

ar»- abb- to make them in field practice, This column was put 



GAINES: SMUT RESISTANCE OF WHEAT. 



221 



Table 2. — Percentage of infected plants, percentage of smut on infected plants, 
and total percentage of loss from smut under conditions of maximum 
infection, together with the total loss from smut under 
conditions of minimum infection. 



Variety. 


Infected plants. 


Smutted heads 
on infected 
piants. 


Total smut 
produced. 


Smut heads in 
field variety 
test. 




Percent. 


Percent. 


Percent. 


Percent. 




26.96 


6.71 


1. 81 


1.8 


Alaska 


I8.60 


14.41 


2. 69 






76.4O 


77.64 


59-32 


3-9 




83.65 


82.01 


68.61 


3-9 




93.20 


75-04 


69.93 




91.27 


86.93 


79-34 


2-5 


Winter Bluestem 


93-24 


87-07 


81.19 


5-6 


Little Club 


87.75 


97-15 


85.14 


8.1 


Jones Winter Fife 


94.17 


93o6 


87.92 


4-7 


Hybrid 143 


93-54 


95-79 


89.60 


9.4 


Hybrid 123 


96.42 


93.28 


89.94 


8-5 


Hybrid 108 


93.62 


96.27 


90. 12 


6-7 


Hybrid 128 


95-io 


96.90 


92.15 


8.6 



in to show that careful seed treatment and crop rotation do not 
eliminate smut, although it is less than is produced by the average 
farmer on summer fallow, according to the estimates of county 
agricultural agents. 

Turkey (Wash. No. 326) is the only highly resistant wheat of 
commercial importance in the list. Even after a plant was infected 
93.29 percent of the normal yield of wheat was produced. In con- 
trast with this, Red Russian, the variety most commonly grown in 
the moister sections of the State, produced only 17.99 percent of the 
normal yield after it was once infected. Moreover, only 26.96 per- 
cent of the Turkey plants were infected, while 83.66 percent of the 
Red Russian plants showed infection. Little Club and Jones Winter 
Fife are much less resistant even than Red Russian. It will be noted 
that 87.75 percent of the plants of Little Club were infected and that 
these produced only 2.85 percent of the normal yield of grain. Jones 
Winter Fife had a higher percentage of infection (94.17 percent), 
but more grain was recovered from the infected plants (6.65 percent). 
The same phenomenon is shown with Turkey and Alaska, the two 
distinctly resistant strains. Turkey has 8.36 percent more infected 
plants, but recovers 7.7 percent more grain from the infection than 
does Alaska. 

Factors Causing Varying Degrees of Resistance. 
From these irregularities it would seem probable that there are 
two distinct factors that control the resistance of wheat to smut. 
One prevents infection, as is seen by the large variation in percent- 



222 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

age of infected plants between Alaska and Hybrid 723. The other 
prevents smut-balls from forming, as is seen by the large variation 
in quantity of wheat produced on infected plants of Turkey and 
Little Club. If these two factors exist, one preventing infection and 
one preventing the smut from fruiting after infection, there is a 
high degree of correlation between them, for in general low percent- 
age of infection is followed with a low percentage of smut-balls pro- 
duced on the infected plants. The converse is also true. High per- 
centage of infection is associated with a high percentage of smut pro- 
duced on the infected plants. 

Little is known concerning the actual process of infection. The 
assumption that all infected plants would produce at least one smut- 
ball may be wrong. We have no means of determining this point. 
Perhaps many infection threads must enter the host cells and there 
fuse in order to live and reproduce smut spores at harvest time. It 
may be, as Kirchner 2 suggests, that a cell sap of slightly higher acid 
content is associated with a high degree of smut resistance. 

Whatever the reason, there is a very marked difference in the 
resistance to smut of some of the different varieties tested. Under 
conditions of maximum infection Turkey was reduced in yield 1.8 
percent by smut, while hybrid 128 was reduced 92.15 percent under 
the same conditions. 

AGRONOMIC AFFAIRS. 

MEMBERSHIP CHANGES. 

The membership reported in the April issue was 648. Since that 
lime (> new members have been added, 4 have resigned, and 1 has died, 
a net ^ain of 1 and a present membership of 649. The names and 
addresses of the new members, names of those resigned and deceased, 
and such changes of address as have been reported are as follows. A 
h'M is also appended of those whose addresses are unknown to the 
Secretary-Treasurer, recent letters sent to them at the latest reported 
addresses having been returned unclaimed. Any one who can furnish 
complete addresses of any of these persons will confer a favor on the 

Treasurer if they will make this information available to 

him. 

2 von Kirchner, 0. In Zeitichr. Pflanzenkrank, 26: 17 25. Stuttgart, Apr. 

S3 ''>''' AVi it .. . (I in Internal. KYv. Sci. Tract. A«r., vol. 7, July, 1916. 



AGRONOMIC AFFAIRS. 



223 



New Members. 

Buenaventura, Linea Rueda, 97 Antiguo Entre 8 y 10, Vedado, Habana, Cuba. 
Cooper, M. L., Merryville, La. 

Covvles, Henry C, University of Chicago, Chicago, 111. 
Garber, R. J., University Farm, St. Paul, Minn. 
Johnson, T. C, Va. Truck Station, Norfolk, Va. 
Kirk, N. M., Bureau of Soils, Washington, D. C. 

Members Resigned. 

Bell, N. Eric, Lathrop, E. C, Miles, Frank C, 

Packard, W. E. 

Member Deceased. 

Anderson, A. C. 

Changes of Address. 
Jones, J. W., Biggs Rice Field Station, Biggs, Cal. 
Krall, J. A., County Agent, Manchester, Iowa. 
Moore, Harvey L., c-o Thos. S. Newell, R. D., Pemberton, N. J. 

Addresses Unknown. 

Bruce, O. C, Douglas, J. P., Kent, W. A., 

Currey, Hiram M., Freeman, Ray, Kenworthy, Chester, 

Shinn, E. H. 



ROLL OF HONOR. 



Since the last issue, the names of several members of the Society 
who are serving their country in its military forces have been reported. 
The complete list reported to date is published herewith. The editor 
will appreciate the favor if those who know of other members of the 
Society whose names should be added will report them to him, as well 
as items of interest regarding any of these men. 

Gray, Samuel D., Quigley, J. V., 

Head, A. F., Ratliffe, Geo. T., 

Holland. B. B., Raymond, L. C, 

Jensen, O. F., Richards, Phil E., 



Brunson, A. M., 
Burnett, Grover, 
Cates, H. R., 
Chapman, James E 
Childs, R. R., 
Downs, E. E., 
Ellison, A. D., 
Gilbert, M. B., 
Graham, E. E., 



Karlstad, C. H., 
Kime, P. H., 
Moo maw, Leroy, 
Palmer, H. Wayne, 

PURINGTON, J. A., 

Towle, R. S. 



schneiderhan, f. j., 
schoonover. w. r., 
Scott, Herschel, 
Smith, John B., 
Tabor, Paul, 



NOTES AND NEWS. 

Carleton R. Ball, for the past several years in charge of work with 
wheat in the western half of the United States in the Office of Cereal 
Investigations, has succeeded M. A. Carleton as chief of that office. 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Recent changes in the Office of Dry-Land Agriculture, U. S. De- 
partment of Agriculture, involve the transfer of J. M. Stephens, su- 
perintendent of the Judith Basin Substation, Moccasin, Mont., to the 
superintendency of the Northern Great Plains Field Station, Mandan, 
X. Dak., vice W. A. Peterson. P. V. Cardon, formerly of the Office 
of Cotton Investigations, has succeeded Mr. Stephens at Moccasin. 

Four field representatives of the Office of Cereal Investigations, U. 
S. Department of Agriculture, resigned recently to engage in farming 
on their own account. With their former positions, they are E. L. 
Adams, superintendent of the Biggs Rice Field Station, Biggs, Cal. ; 
L. R. Briethaupt, superintendent of the field station at Burns, Oregon ; 
N. C. Donaldson, of the Judith Basin Substation, Moccasin, Mont. ; 
and J. D. Morrison, of the Highmore substation, Highmore, S. Dak. 
Mr. Adams has been succeeded at Biggs by J. W. Jones, former 
superintendent of the Nephi (Utah) substation, who in turn has been 
succeeded by A. F. Bracken. J. H. Martin, formerly of the Belle- 
fourche Experiment Farm, Newell, S. Dak., is the new superintendent 
at Burns, Ore. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. io. September, 1918. No. 6 



THE TRIANGLE SYSTEM FOR FERTILIZER EXPERIMENTS. 1 

Oswald Schreixer and J. J. Skinner. 
Introduction. 

Fertilizer experimentation for determining the specific needs of any 
particular soil type or crop is one of the big problems before Amer- 
ican agriculturists. It is not our purpose here to dwell upon the 
shortcomings of many efforts in this direction, but we must say in 
passing that the popular conception, even among agricultural spe- 
cialists, that this problem can be solved by a soil, or a plant, or an ash 
analysis is a vain hope which has not and cannot be realized. Much 
can be learned from such work, but not the fertilizer requirement of 
the soil or plant to increase the yield, quality, appearance, or freedom 
from disease. Experimentation direct with soil and plant have thus 
far been the only means to give this answer and in this connection the 
soil has nearly always been ignored and the fertilizer combinations 
tested have always been so restricted that a full and complete answer 
to this complicated question is yet to be reached. There have been 
some excellent fertilizer experiments, especially the long-term systems 
at several of the experiment stations, but by far the greater number 
of tests made from time to time on this land or that land, this crop or 
that crop, the country over, have been so lacking in plan and in thor- 
oughness that they have served only a temporary purpose. How- 

1 Contribution from the Office of Soil Fertility Investigations, Bureau of 
Plant Industry, U. S. Department of Agriculture, Washington, D. C. Pre- 
sented by the senior author, with illustrations, at the tenth annual meeting of 
the American Society of Agronomy, Washington, D. C, November 13, 1917 
Received for publication April 3. 1918. , 

225 



226 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



ever useful they may have been, they offer little for the interpreta- 
tion of the broad fundamental questions of soil fertilization and crop 
needs. It is not our purpose to disparage the many well-planned ex- 
periments which are in successful operation. Rather, we wish to call 
to the attention of those planning further work, a system of fertilizer 
experimentation which, with proper and careful attention to soil dif- 
ferences as far as they can be mapped in advance or in the course of 
the experiment, ought to give a sufficiently comprehensive basis for 
proper interpretation and easy presentation and handling of the 
results. 

In the following discussion the triangle system is outlined as we 
have used it in our problems and as others have employed it in similar 
or related lines, largely as "a suggestion to other workers who may not 
as yet have seen its advantages. 

When it is desired to test the effect of all possible ratios of the fer- 
tilizer elements, PoO-, NLL, and K 2 0, the triangle system has been 
found to be admirably suited. Graded in 10 percent stages there 
are in this experiment 66 tests, or graded in 20 percent stages there 
are 21 tests, involving the fertilizer elements singly, in combinations 
of two, and in combinations of three. 

To bear in mind these various ratios and the results obtained there- 
with is difficult. In so comprehensive an experiment as this the ma- 
terial must be reduced to a workable basis so that the various phases 
of the results can readily be presented and the proper correlations and 
comparisons made. 

Use of the Triangular Diagram. 

A triangular diagram is shown in figure 28. It is an equilateral 
triangle in which the extreme points of the angles represent 100 per- 
cent- respectively, of the constituents, P 2 5 , NH 3 , and K 2 0, as shown 
in the diagram. Each side of the triangle is divided into ten equal 
part£ and lines are drawn connecting these points. :i 

In the diagram, for the sake of ready reference, the intersections 
of tfiese lines have been numbered. I f we consider the line repre- 
senting ili<- base of the triangle, it is obvious that the point which rep- 
resentl (00 percenl K ,( I 150 in the diagram) represents at the same 
time n prrrcnt NIL, and the point which represents 100 percent NIL. 

- The t ' tin " i'H) percent," ;is lu re used, refers to the maximum quantity of 
a liflgle fertilizer constituent chosen for the experiment 

Sttdl diagrams for physical < hemical work, ^ivin^ still finer rulings, namely, 
100 to e;i< h hue, c ;m l.e purchaseo" from the Cornell Cooperative Society, 

Ithaca, N. Y. 



SCHREINER & SKINNER: FERTILIZER EXPERIMENTS. 22J 



P 2 s 



(66) likewise represents o percent K 2 0. If we take a point half 
way between these two points (61) we have a mixture of the two 
salts in equal proportions ; i. e., a mixture of the salts represented by 
that point will be 50 percent K 2 and 50 percent NH 3 . Similarly, 
point 16 represents 50 percent K 2 and 50 percent P 2 O s , and point 
21 represents 50 percent NH 3 and 50 percent P 2 5 . 

If we take a 
point nearer to 
either of the cor- 
ners, we will have 
a higher percent- 
age of one and 
a correspondingly 
lower percentage 
of the other. For 
instance, at point 
59 the composi- 
tion is 70 percent 
K 2 and 30 per- 
cent NH 3 ; at 29 
it is likewise 70 
percent K 2 0, but 
30 percent P 2 O s ; 
at 64 it is 20 per- 
cent K 2 and 80 
percent NH 3 ; at 
45 it is likewise 80 percent NH 3 but 20 percent P 2 5 . 

As stated above, points on the base line 56-66 represent mixtures 
containing no P 2 5 . The next line above this, namely 46-55, repre- 
sents mixtures containing throughout 10 percent P 2 5 , but varying 
amounts of the other two constituents. Similarly the line 37-45 rep- 
resents throughout 20 percent mixtures of P 2 O s ; line 29-36, 30 per- 
cent mixtures of P 2 5 , and so on upward until point 1, the apex of 
the triangle, is reached, where the composition is 100 percent P 2 5 as 
already explained. Similarly, points on the line 1-66 represent o 
percent K 2 ; line 2-65 represents 10 percent K 2 but varying 
amounts of P 2 5 and NH 3 , and so on until at point 56 the composi- 
tion is 100 percent K 2 0. Likewise points on the line 1-56 represent 
o percent NH 3 ; line 3-57 represents 10 percent NH 3 , but varying 
amounts of P 2 O s and K 2 0, and so on until at point 66 the composi- 
tion is 100 percent NH 3 . It is therefore obvious that any point 




k 2 o NH 3 
Fig. 28. Triangular diagram with the points numbered, 
representing the 66 fertilizer combinations in 10 percent 
stages. 



228 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



within the triangle represents a 100 percent mixture composed of 
three constituents, its position in the triangle being determined by the 
composition. For instance, point 12, being on the 60 percent phos- 
phate line represents that composition of P 2 O s , namely 60 percent, 
and being at the same time on the 10 percent NH 3 line and the 30 
percent K 2 line it represents 10 percent and 30 percent of these 
constituents respectively. The composition of the mixture repre- 
sented by this point is therefore P 2 O s 60 percent, NH 3 10 percent, 
and KoO 30 percent, i. e., the composition of the fertilizer mixture is 
60-10-30. Similarly the point 34 represents a mixture of the com- 
position P 2 O s 30 percent, NH 3 50 percent, K 2 20 percent, or a fer- 
tilizer composition of 30-50-20. 

It is of course 
evident that any 
other percentage 
composition or ratio 
of fertilizers could 
likewise . be repre- 
sented on such a dia- 
gram. Commercial 
brands can be rep- 
resented on such a 
diagram by reducing 
the sum of the com- 
mercial percents to 
as//, a basis of 100, thus 

Fig. 29. Triangular diagram with the points num- an ^4~4 fertilizer 
bered, representing the 21 fertilizer combinations in becomes $0—25-25. 
20 percent stages. This point lies half- 

way between 18 and 

i';. Similarly a jo 2-2 fertilizer becomes 72-14-14 and lies between 
8 and 9 on the diagram. For accurately locating such points, the 
finely ruled and subdivided paper referred to in the footnote should be 
used. 

Jn stating the percentage composition of the fertilizer mixtures 
in such work, the figures are always given in the order — PzOb, NII 3 , 
and K.O ;i Oiown above. The symbols V.X) r , t NIT.,, and K 2 are 

u ed in conformity with fertilizer practice, even when the nitrogen, 

for instance, is in the nitrate form. 

The triangle therefore represents single fertilizer constituents at 
pices or vertices, mixtures of any two constituents along the 




SCHREINER & SKINNER: FERTILIZER EXPERIMENTS. 229 

boundary lines of the triangle, and mixtures of all three constituents 
within the triangle. 

In Table I is given the composition represented by each of the 
66 points in the diagram. 



Table i. — Sixty-six possible ratios of the three fertilizer constituents, 
P2O5, NH 3 , and K 2 in 10-percent stages. 



Point 
No. 


Ratio or percentage 
composition. 


Point* 
No. 


Ratio or percentage 
composition. 


Point 
No. 


Ratio or percentage 
composition. 


P2O5. 


NH 3 . 


K 2 0. 


P2O5. 


NH 3 . 


K 2 0. 


p 2 o 5 . 


NH 3 . 


K 2 o. 


I 


IOO 








23 


40 


10 


50 


45 


20 


80 





2 


90 





10 


24 


40 


20 


40 


46 


10 





90 


3 


90 


10 





25 


40 


30 


30 


47 


10 


10 


80 


4 


80 





20 


26 


40 


40 


20 


48 


10 


20 


70 


5 


80 


10 


10 


27 


40 


50 


10 


49 


10 


30 


60 


6 


80 


20 





28 


40 


60 





^o 


10 


40 


50 


7 


70 





30 


29 


30 


O 


70 


5i 


10 


50 


40 


8 


70 


10 


20 


30 


30 


10 


60 


52 


10 


60 


30 


9 


70 


20 


10 


31 


30 


20 


50 


53 


10 


70 


20 


10 


70 


30 





32 


30 


30 


40 


54 


10 


80 


10 


11 


60 





40 


33 


30 


40 


30 


55 


10 


90 


O 


12 


60 


10 


30 


34 


30 


50 


20 


56 


O 





IOO 


13 


60 


20 


20 


35 


30 


60 


10 


57 





10 


90 


14 


60 


30 


10 


36 


30 


70 





58 





20 


80 


15 


60 


40 





37 


20 





80 


59 





30 


70 


16 


50 





50 


38 


20 


10 


70 


60 





40 


60 


17 


So 


10 


40 


39 


20 


20 


60 


61 





50 


50 


18 


50 


20 


30 


40 


20 


30 


50 


62 





60 


40 


19 


50 


30 


20 


4i 


20 


40 


40 


63 





70 


30 


20 


50 


40 


10 


42 


20 


50 


30 


64 





80 


20 


21 


50 


50 





43 


20 


60 


20 


65 





90 


IO 


22 


40 





60 


44 


20 


70 


10 


66 





IOO 


O 



In Plate 5 the composition of the 66 ratios varying in 10-percent 
stages are shown as circles with red, black, and white segments to 
represent visually the proportions of P 2 5 , NH 3 , and K 2 0. 

Preparation of the Fertilizer Mixtures Varying in io-Percent Stages. 

In order to make clear the manner of preparing fertilizer mixtures 
for such an experiment, let us assume that a test of acid phosphate, 
sodium nitrate, and potassium chloride is to be made, that the plats 
are to be 100 square feet, and that the fertilizers are to be applied 
at the rate of 50 pounds per acre of the active fertilizer constituents. 
This means that the sum total of P 2 5 , NH 3 , and K 2 will be in all 
cases 50 pounds per acre. As the size of the plats in these experi- 
ments is 100 square feet, it becomes necessary to calculate the quan- 
tity of each fertilizer required in the total of 66 fertilizer combina- 
tions so as to obtain the proper quantity and proper ratio for each plat. 



23O JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Suitable containers are used for holding these fertilizer combina- 
tions until required for use in the field. These are numbered from 1 
to 66, corresponding with the numbers of the plats in the experi- 
ment. These containers, for a small experiment like this, are set in 
the form of a triangle, as shown in the diagram of Plate 5. The 
addition of the respective analyzed fertilizer substances is thus made 
as follows : 

Acid phosphate. — If the acid phosphate analyzed 14.0 percent P 2 O s 
and the application is to be at the rate of 50 pounds of P 2 O s per 
acre, then 357.15 pounds of the acid phosphate will be required. For 
a plat of 100 square feet, 0.8175 pound or 370.3 grams will be re- 

Table 2. — Grams of acid phosphate (14 percent P 2 5 ) required in different 

fertiliser mixtures. 



Fertilizer container along line 
in Plate 5. 


Quantity of acid 
phosphate, 14 per- 
cent P2O5. 


Fertilizer container along line 
in Plate 5. 


Quantity of acid 
phosphate, 14 per- 
cent P2O5. 


I 


Grams. 
370.3 
333-3 
296.2 

259-3 
222.2 


22-28 


Grams. 

148. 1 

in. 1 
74-1 
37-0 

none 


2-3 


29-36 


4-6 


37-45 


7-io 


46-55 




55-66 



quired. This, then, is the quantity to be put into container No. I. 
The next row of containers, Nos. 2 and 3, is to have only 90 percent 
of this quantity (45 lbs. P 2 5 per acre), namely, 333.3 grams. The 
next row of containers is to have only 80 percent of the full quantity 
('40 11)-. VS)r per acre), namely, 296.2 grams. This gradation of 



Table 3. — Grams of sodium nitrate (16.3 percent NHzJ required in different 

fertilizer mixtures. 



Quantity of sodium 
Fc.t.l./r, , ootainw along l.ne ni(rate l6 



Fertilizer container along line 
in Plate 5. 



Quantity oi sodium 
nitiate, 16.3 per- 
cent N H|. 



(,(> . . . 
55-65 
I" <>-\ 

28-62 . 
21 f,i . 



Grams. 
319.O 
287.1 
255-2 
2233 
191.4 



1 5 (><> 
10-59 
6-58 
3-57 
1-56 



(irii)>i\. 
127.6 
95-7 
63.8 

31.9 
none. 



the quantity of acid phosphate which is to he weighed out and put 
into tin respective containers is host shown by the figures in Table 2. 
Sodium nitrate. If the ^odium nitrate analyzes 16.3 percent NfT 3 , 



SCHREIXER & SKINNER: FERTILIZER EXPERIMENTS. 



231 



306.7 pounds will be required to have 50 pounds NH 3 per acre. For 
the plat of 100 square feet this will be 0.7041 pound or 319.0 grams. 
Container No. 66 will therefore receive this full quantity. Con- 
tainers Nos. 55 and 65 receive only 90 percent of this, 287.1 grams, 
and so on for other lines of containers according to the figures in 
Table 3. 

Table 4. — Grams of potassium chloride (51 percent K 2 0) required in different 

fertilizer mixtures. 



Fertilizer container along 
line in Plate 5. 



Quantity of 
potassium chloride 
51 percent K2O. 


Fertilizer container along 
line in Plate 5. 


Quantity of 
potassium chloride 
51 percent KoO. 


Grams. 




Grams. 


101.9 


1 1-62 


40.8 


91.7 


7-63 


30.6 


81.5 


4-64 


20.4 


71-3 


2-65 


10.2 


61. 1 


1-66 


none. 


50.9 







56... 

46-57 
37-58 
29-59 
22-60 

16-61 



Potassium chloride. — If the potassium chloride analyzed 51 percent 
K 2 0, then 98 pounds of chloride are required for an application of 
50 pounds of K 2 0. For the plat of 100 square feet this is 0.2249 
pound or 101.9 grams. The various containers will receive the quan- 
tities specified in Table 4. 

Table 5. — Quantity of fertilizers to be applied per acre at the rate of 100 pounds 
of the sum of P 2 5 , NH 3 , and K 2 0, tn 20-percent stages. 

Fertilizer Pounds of acid phosphate, Pounds of sodium nitrate, Pounds of potassium sulfate, 
No. 16 percent P2O5. 19 percent NH3. 50 percent KoO. 



I 


625 










2 


500 







40 


3 


500 


105 







4 


375 







80 


5 


375 


105 




40 


6 


• 375 


210 







7 


250 







120 


8 


250 


105 




80 


9 


250 


210 




40 


10 


250 


316 







" 


125 







160 


12 


125 


105 




120 


13 


125 


210 




- 80 


14 


125 


3i6 




40 


15 


125 


421 







16 










200 


17 





105 


« 


160 


18 





210 




120 


19 





3i6 




80 


20 





421 




40 


21 





526 








232 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Preparation of the Fertilizer Mixtures Varying in 20-Percent Stages. 

For larger experiments in the field a smaller number of plats may 
be considered more practical. For this purpose the variations are 
placed at 20 percent, as shown in figure 29. Otherwise the system 
is the same as that described for the 10-percent stages. The manner 
of preparing the fertilizer mixtures and the character of the con- 
tainer naturally varies with the size of the plats to be treated. Table 
5 shows these 21 mixtures based on the usual composition of the com- 
mercial fertilizer material used. A similar table is readily constructed 
for any other analysis. The table is put on an acre basis ; if the plat 
is one-tenth acre in size, one-tenth of the quantity is taken; if one- 
twentieth, one-twentieth of the quantity is taken, etc. The table is 
likewise based on an application of 100 pounds of the active fer- 
tilizer constituents. If 50 pounds are to be applied half of the quan- 
tity is taken, if 200 pounds is to be applied, twice the quantity in the 
table is taken, and so on. 

The Triangle System in Nutrient Solution Studies. 

In nutrient solution work the triangle has been found to be very 
useful, and it was in the successful prosecution of this work that the 
system originated. 4 The experiments comprised the 66 cultures men- 
tioned, using the pure, soluble salts CaH 4 (P0 4 ) 2 , NaNO s , and K 2 S0 4 
in the proportions required by the system and with a total concentra- 
tion in each culture of 80 parts per million of P 2 O r „ NH 3 , and K 2 0. 
Figure 30 represents the green weight of wheat plants in grams ob- 
tained in such an experiment, recorded at the proper place in the dia- 
gram according to the nutrient solution in which it was grown. This 
is shown to illustrate the orderly and concise way in which the 66 dif- 
ferent experiments are unified and presented to the experimenter for 
intelligent interpretation. A graphic representation of the numbers 
in figure 30 is given in figure 31. In this diagram the^ area of the 
circle is proportional to the numbers of figure 30 and is made by find- 
ing the radius corresponding to these numbers as areas of a circle ac- 
cording to the formula: A= — — for which R can either be calcu- 

4 

lated Of taken from a table of such values to be found in many books. 
Notice how the numerous results present the story of the experiment 
a a whole, the relative values of the individual salts, the relative values 
of tin- two constituents as shown by the outlines of the triangle, and 

*Schr< ner, ( > vrald, and Skinner, J. J. Ratio of phosphate, nitrate, and 

pota mm on absorption ami growth. /;; Hot. da/.., 50: 1-30. 1910. 



Journal of the American Society of Agronomy. 



Plate 5 




Plate 5. Triangular diagram showing the 66 fertilizer mixtures in [0 perceni 
.1 variation, ;t> indicated by the colors. Mack, white and red. 



SCHREINER & SKINNER ', FERTILIZER EXPERIMENTS. 



233 



the relative values of the three constituents combined. Within the 
latter group it is at once apparent that the largest growth is in the 
middle of the lower half of the triangle as a whole. This is the region 
of greatest growth and these are the culture solutions in which the 
ratios are best suited for plant development. 

A still further illustration of the usefulness of such a diagram be- 
comes apparent when the composition of the crop grown, the ash con- 
stituents, protein content, starch content, or any other analytical figure 

p i°5 



.6 48 




Fig. 30. Diagram showing method of recording the results in figures at the 
proper place on the diagram. Wheat in solution culture. 

such as alkaloidal content, etc., is considered in relation to fertilizer 
practice or influence. For instance, in the above nutrient culture ex- 
periment, the 66 ratios of P 2 5 , NH 3 , and K 2 were analyzed after 
the plants had grown in them for a period of days so as to determine 
the changes which had taken place in the ratio of the constituents and 
thus arrive at the ratio which the plants had absorbed. For the three 
constituents this amounted to 198 determinations and with the orig- 
inal 198 figures and the 198 figures giving the absorbed constituents 
make a total of 594 individual figures. To obtain a clear conception 
of what was going on by an interpretation of the analytical results 



234 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



would tax the mind of anyone to distraction. Here the triangular 
diagram serves a useful and necessary purpose. In figure 32 the black 
dots give the original composition in P 2 O s , NH 3 , and K 2 0, the small 
/g . circles the compo- 

O sition of the solu- 

^ tion after growth, 

the arrow points 
o O O the composition of 

q Q Q nutrients removed 

n n n o n by the growing 

° UUU ^ plant. Notice the 

O (^) (^) (^) O unity presented by 

n /~\ /~\ /*-\ ^ this representation 

Acr^r^r^r^\ ° f theSe multitudi " 

O OO O O O O O nous and otherwise 

OOOOOOOO unintelligible re- 

0000000000 ;f ; j h t e rr f - 

OOOOOOOOOOO inite region, that is, 

the plants, no mat- 

FiG. 31. Diagram showing method of recording the ter j n w h a t solution 

results diagrammatically as areas of circles at the . 

. « «• r rr-»i they were growing, 

proper points in the diagram same as figure 30. The J & 

largest circles indicate the regions of greatest growth. a ^ l eas t attempted 

to get the composi- 
tion best suited for their development and this lies in the middle of 
the triangle, corresponding with the area of greatest growth shown 
in figures 30 and 31. 

In our experiments, as is well known, the principal purpose was a 
study of the influence of certain toxic organic compounds on the 
growth and on nutrition. For this purpose a second set of 66 cultures 
is treated with a certain definite quantity of the substance in each cul- 
ture, and comparison with the corresponding normal culture is then 
made. In this way the influence of the substance on character and 
extenl of growth, and influence of the fertilizer in overcoming tox- 
icity ifl then Studied, Such influence is shown in figure 33, wherein 
the area of greatest growth is shown diagrammatically as it occurred 
under the influence of different toxic substances, shifting this from 
the normal in one direction Or the Other. 8 The significant fact is that 

■"• Schreiner, Oswald, and Skinner, |. J. Sonic effects of a harmful organic 
soil constituent. In hot. i>a/., 50: 161-181. 1910. 

Schreiner, irald, and Skinner, J. J. The toxic action of organic compounds 

as modified \>y fertilizer salts. In hot. ( 54: 31 4H. 1912. 



SCHREIXER & SKIXXER: FERTILIZER EXPERIMENTS. 



235 



while the region of greatest growth in the case of cumarin is displaced 
toward the higher phosphate region of the triangle, this is also the 
fertilizer constituent which more than any other is antitoxic to cu- 
marin ; similarly, potash appears antitoxic to quinone and nitrate to 
vanillin. 

These characteristics are brought out by a system of grouping the 
individual culture results, according to those which contain, for in- 

P Z Q S 




^2° NH 3 

Fig. 32. Diagram showing the ratio of the original, the final, and the ratio 
of the loss of P 2 5 , XH 3 , and K 2 from the culture solution. The dots indi- 
cate the ratio of the constituents in the original solution; the circles show the 
ratio of the constituents in the solution after growth; and the arrows show 
the ratio of the decrease. 



stance, more than 50 percent phosphate, or more than 50 percent 
nitrate, or more than 50 percent potash, as shown respectively by the 
dotted lines in figure 34, where the results for cumarin are presented. 
The cultures included as mainly phosphatic are best seen by reference 
to figure 28 or plate 5. They are those comprised in the sub-triangle 
1-16-21. Similarly, the mainly nitrogenous are those included in the 
sub-triangle 66-21-61, and the mainly potassic those included in the 
sub-triangle 56-61-16. From figure 34 it is apparent that in the 



236 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 




mainly phdsphatic cultures the cumarin is without effect or that the 
phosphate has had an antitoxic affect. 

p 2 o 5 There is still another way 

of handling the data which 
confirms conclusions reach- 
ed in this manner, or brings 
out new relationships and 
that is to add up the cul- 
tures along lines and obtain 
an average result for each 
line. For instance, in figure 
35, the bottom line contains 
no phosphate, while the 
next lines contain consecu- 
tively 10, 20, 30, 40, 50, 60, 

NH3 70, 80, 90 per cents, and 100 
Fig. 33. Diagram showing method of rep- 1 1 ~, 

resenting the region of greatest growth under P ercent phosphate. 1 he 
different conditions. relative growths along these 

various phosphate lines in 
the case of cumarin arrange themselves in an orderly fashion with in- 
crease in phosphate. 

Harris has used _ . 

our triangular sys- 
tem in studying 
the effect of alkali 
salts on the ger- 
mination and the 
growth of jjiants. 
Harris used the 
salts in 25-percent 
stages as shown in 
figure 36, which is 
taken from his ar- 
ticle. I'esides the 
combination so- 
d i um-chloride-sul- 

fate-carbonate he 

•CO NH 3 

used a number of T4 . . 4 , a . , 

rio. 34 Diagram snowing the method ox grouping 

Other SaltB and COm- culturei according to the prevailing fertilizer element. 

binatiom I be u-' 

* Harris, ]•', S. Effect of alkali salts in soils on the germination and growth 
of crops. In Jour. Agr. Research, v. 5, no. I, p. 1-53. 10] 5, 




SCHREINER & SKINNER: FERTILIZER EXPERIMENTS. 



237 



of the system in 25-percent stages has one rather serious drawback in 
certain work on account of the fact that the combinations of all three 
constituents together 
are rather limited, in 
fact confined to three 
cultures. For the pur- 
pose used this may 
have been permissible, 
but for a fertilizer or 
nutrition study these 
25-percent variations, 
in our opinion, are 
not as suitable as the 
20-percent . stages, 
which give six com- 
binations containing 
the three constituents. 
Figure 36 illustrates 
the composition of the 
solutions of salt mix- 




Fig. 35. Diagram showing the method of group- 
ing according to lines of cultures containing a con- 
stant amount of any one fertilizer element. Shows 
tures by shading in- tne influence of phosphate in overcoming the toxic 
stead of by colors as effect of cumarin. 
in Plate 5. 

Harris further used the diagram in representing his results in a 
way which is worthy of further consideration and is shown in figure 
37, taken from his article. He represents his cultures by small circles 
arranged in the triangle form according to the composition. In the 
circle dots represent the number of plants germinated, and a dash 
the amount of dry matter produced. The diagram therefore shows 
at a glance the composition of the salt mixtures, the germination ob- 
tained, and the dry weight of the crop for any one concentration. 
Each concentration of salts is represented by another diagram and 
the series gives a very intelligent and comprehensive view of the re- 
sults he obtained in his experiment. In the particular illustration 
here shown it becomes readily apparent that with increasing concen- 
tration, germination is interfered with, and that this shows itself first 
in the (NH 4 ) 2 CO s section, and later also in the K 2 CO s section, with 
the least effect in the Na 2 C0 3 section. Similar results are apparent 
in the growth line in the series of diagrams. This is a striking illus- 
tration of the varied usefulness of having a definite fundamental plan 
in the experiments which permits handling complicated results in a 



2 3 8 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



simple manner, both for the purpose of experimentation, recording 
the results, and presenting them in a clear, concise, and logical way 
to others. 

This triangular system has been used and amplified by Tottingham, 7 
while working at the Johns Hopkins University, in his rather com- 




I i' . 36. Diagram showing percentage of salts, mixtures, and their position 
in the diagrams of experimental sets, as used by Harris. Variations in 25- 
pcrccnt stages. 



prehensive physiological study of nutrient solutions, for the details 
of which the reader ifl referred to the original article. Likewise 
Shive, 1 in following up this subject at the same institution, used the 

r Tottingham, \V. K. A quantitative chemical and physiological study of 
nutrient ■-'.hition for plant cultures. /// Physiological Researches, vol. I, no. 4, 
P. U3-245 1914 

* Shivc, J. W. A study of physiological balance in nutrient media. In 
Physiological Researches, vol. 1, no. 7, p. 327-397. 1915. 



SCHREINER & SKINNER: FERTILIZER EXPERIMENTS. 



239 



triangular system with good effect. Both of these workers omitted 
the outside lines of the triangle, confining themselves to the more com- 
plete mixtures represented by the interior of our triangle system. 
Both investigators made up their solutions on the basis of osmotic 
pressure instead of percentages. The salts used by Tottingham 
were Ca(N0 3 ) 2 , KNO aj KH 2 P0 4 , MgS0 4 , while Shive omitted the 
KNO3 as contributing no ions not supplied by the other salts. More 
recently WolkofP at Rutgers College, in continuing the work of Shive, 
substituted (NH 4 ) 2 S0 4 for the KNO s in Tottingham's solution, using 
the triangle for study and interpretation. For the results obtained 
by these workers the originals must be consulted. 




cocoppm zopopptn e.oooppirt zcooppm. to.oooppm 

. = One plant. — —0.1 gm. dry matter. 



Fig. 37. Method of representing results as used by Harris. Diagram show- 
ing the number of wheat plants up and dry matter produced in 24 days on 
Greenville loam with ammonium carbonate, sodium carbonate, and potassium 
carbonate in different combinations and concentrations. 

McCall 10 has also used our triangle system with interesting results 
in his work on the physiological balance of nutrient solutions. The 
salts used were calcium nitrate, magnesium sulfate, and potassium 
acid phosphate. The composition of his solutions used in sand cul- 
tures are represented by the diagram reproduced in figure 38. The 
variations are in 10-percent stages, but McCall, like Tottingham and 
Shive, omits the outside lines which comprise the constituents alone 
and the combinations of two, confining himself to a consideration of 
the variously constituted mixtures of the three components. 

9 Wolkoff, M. I. Effect of ammonium sulphate in nutrient solution on the 
growth of soybeans in sand cultures. In Soil Science, 5: 123-150. 1918. 

10 McCall, A. G. The physiological balance of nutrient solutions for plants 
in sand cultures. In Soil Science, 2 : 207-253. 1916. 



240 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



True and Bartlett, 11 in their study of the exchange of ions by plants 
growing in solution cultures, used this system. The salts used were 
the nitrates of potassium, calcium, and magnesium. The composition 
is stated in terms of concentration as to normality instead of on a 
percentage basis. The plan is reproduced in figure 39. Each point 




Fig. Triangular diagram used by McCall, showing the arrangement of 
the sand cultures with respect to the partial concentrations of the three salts 
employed. Unshaded segments represent the proportions of Ca(N0 3 ) 2 : stip- 
pled tegmenta the MgSO«: and the segments shaded with crosses the KILPO,. 
The best nine cultures are marked X, while the poorest are marked 0. 



in the figure represents a solution, the original composition of which 
i- indicated on the three intersecting lines reading upward from the 
inter <•« lion. The stun of the numerals at the intersection at any point 
is i.jo, indicating a total concentration of 140N X 10 ". The figure! 

given at the intersections for the residual concentration at the time 

11 'I mi', R II . and Bartlett, H, II. The exchange of ions between the roots 
of Luf u albus and culture solutions containing three nutrient salts. In 
Amcr. Jour. Botany, 3: 47-57. 1016. 



SCHREINER & SKINNER: FERTILIZER EXPERIMENTS. 



241 



of maximum absorption are based on the original concentration, in 
each case, 140^ X icr 6 being considered as 1.000. 

True 12 also used the triangle system in his study of the effects of 
calcium in its relation to plant nutrition. 




Fig. 39. Triangular diagram as used by True and Bartlett, in showing the 
residual concentration of solutions containing KN0 3 , Ca(N0 3 ) 2 , and 
Mg(N0 3 ) 2 , at the time of maximum absorption. 



Chamot 13 used the triangular diagram in connection with a study 
of certain media employed for the bacteriological examination of 
water. In using the triangular diagram there were considered (1) 
the concentration of peptone, (2) the concentration of the inorganic 
salts present, and (3) the nature of the reaction of the medium. He 

12 True, R. H. Calcium in its relation to plant nutrition. Presented before 
the American Society of Agronomy at its Washington meeting, 1917. Un- 
published. 

13 Chamot, E. M., and Redfield, A. W. I. The Schardinger-Dunham medium 
for testing for the presence of hydrogen sulphide forming bacteria. In Jour. 
Amer. Chem. Soc, 37: 1606-1630. 1915. 

Chamot, E. M., and Sherwood, C. M. II. Lactose-peptone media. In Jour. 
Chem. Soc, 37: 1949-1959. 1915. 



242 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



also applied it to mixtures of four components, i. e., in addition to 
those noted above he used carbohydrates, but in this case he followed 
the procedure already mentioned in connection with the use of toxic 
organic substances earlier in this paper, viz., by keeping this fourth 
component constant in all the cultures of the triangle. 

The Triangular System in Field Experiments. 
Following the use of the triangle system in our solution culture 
work, the same system was used in field studies on the action of fer- 
tilizers. The system was first laid out on the Arlington Farm in 
1909, and at the Pennsylvania Agricultural Experiment Station Farm 



1 , 


3 


6 


10 


1 15 


B 


E 


36 


45 


E 


66 ' 


Check 




| Chaok 


B 


n 




9 


14 


L) 


□ 


B 


44 


54 


!•• 






i 


B 


E 


4 




13 


19 


□ 


□ 


B 


E 


E 


| Check | 




| Cheek 


56 


47 | 


S7 




E 


16 


IS 


□ 


E 


B 


B 








E 


46 


36 


29 


F 


Fl 


24 


Fl 


R 


si 


B 







Cheek 


E 


49 


39 


30 


22 


M 


28 


31 


□ 


□ 


«i 








61 


50 


40 


31 


23 


16 


22 


30 


B 


I- 


60 | 


Check 




| Check | 


B 


SI 


41 


E 


F 


17 


11 


21 J 


31 | 


H 


Lil 








E 


E 


42 


| 33 


B 


18 




E 


"1 


47 | 


□ 


B 




Check 


64 


B 


E 


34 


26 


E 


13 


6 


4 


-I 


H 








SB 


44 


B 


E 


27 


20 


B 


9 


5 


2 


Id 


Ed 




H 


l <6 l 


B 


a 


B 


l Z * 


l"l 




B 




3 | 


B 



I J', 40. I Mat showing the field arrangement of two triangle systems in small 

rectangular plats. 



in 1 <)i<>. The plats used arc small and therefore arranged in the 
form of the triangle, although tliis is by no means a requisite of the 
system. The checks run across the field diagonally at intervals and 
a Itrip of original Untreated Boil is maintained in sod between the 
triangles. 

At tlx- Pennsylvania Agricultural Kxperiment Station Farm the 
fertilizer tetl made on ^rass land. When the experiment was 



SCHREINER & SKINNER: FERTILIZER EXPERIMENTS. 243 

begun in 1910, the area on which the plats were located had a fairly 
uniform stand of grass consisting almost entirely of Canada blue- 
grass, Kentucky bluegrass, and timothy, with a very little white and 
red clover. It was estimated that Canada bluegrass covered about 50 
percent of the area, Kentucky bluegrass about 30 percent, and timothy 
about 20 percent. 

The soil is of the Hagerstown series. The surface soil varies in 
depth from 7 to 11 inches, and mechanical analyses of the soil and 
subsoil of plats taken at regular intervals show that the surface soil 
is silt loam and silty clay loam and the subsoil, clay and silty clay 
loam. 

The plats were laid off according to the triangular system as shown 
in figure 40. They are 10 feet square and are separated by 2-foot 
paths. The plats were laid off in duplicate with a 10-foot space 
separating the two series, or triangles. Besides the 66 treated plats, 
there are 6 check plats in each series. The unused dividing strip and 
the 2-foot paths are kept in grass which is cut when the grass on the 
plats is cut. Here one may study the untreated soil. 

As laid out, the two series form a rectangle, and all outside corners 
of the plats on the boundaries are marked with posts 4X4 inches 
which extend out of the ground 12 inches. These are outside of the 
plat so that the inside corner of the post coincides with the outside 
corner of the plat. For any desired purpose the boundaries of the 
plats are defined by stretching twine both ways across the two tri- 
angles. 

The fertilizers have been applied in early April of each year just 
after the grass began to grow. The fertilizers are weighed out into 
wide-mouth bottles as previously described. After marking off the 
plats, the content of each bottle is mixed with about a quart of dry 
sand and then distributed by hand as evenly as possible. The fer- 
tilizer which each plat receives is at the rate of 50 pounds per acre of 
the active fertilizer constituents. 

Some of the results obtained in this grass experiment have been 
presented in an earlier report, 14 but there are further interesting 
changes in the composition of the grasses which will form the subject 
of a later report. These changes in grass composition appear to re- 
sult from the different fertilizer ratios and we wish here to call par- 
ticular attention to the use of the triangle system and diagram in 

14 Noll, C. F., Schreiner, Oswald, and Skinner, J. J. Fertilizer ratio experi- 
ments with grass on Hagerstown loam. In Ann. Rept. Pa. State Coll. for the 
year 1913-1914, pt. 2, p. 22-36. 1915. 



244 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



bringing out these facts. Each plat was analyzed botanically and the 
proportion of Kentucky bluegrass, Canada bluegrass, timothy, red 
clover, and white clover present stated on a percentage basis. This 
means 330 figures for each triangle and these are very easily handled 
by representing each plat or point in the triangle by a circle and letting 
each grass percentage represent a segment of the circle, much like the 
diagram of Plate 5, using 3 different shades of green for the three 
grasses and two shades of red for the clovers. When this is done it 
becomes strikingly apparent that the clovers are decidedly more prom- 
inent in that area of the triangle which is low in nitrogen. Originally 
Canada bluegrass was predominant. In the higher nitrogen area, 
Kentucky bluegrass is now predominant and the Canada bluegrass 
appears to be gradually crowded out. TJiese facts could not have 
been easily seen or portrayed without the use of the triangular sys- 
tem of experimentation and the use of the diagram in the interpreta- 
tion of the results. 

At the Arlington Farm the manner of laying out the plats and the 
size are the same as the grass plats just described, but there are four 
triangles in the experiment. One of these is devoted to the growing 
of wheat, year after year, while the other three are devoted to a rota- 
tion of wheat, corn, and cowpeas, each crop being grown on one of 
the triangles. Plate 6 shows two general views of the field devoted to 
this work. Plate 6, figure I, shows the preparation of the ground for 
planting after the fertilizers have been applied, the stakes marking 
the individual plats, while figure 2 shows the harvested crop from the 
plats. 

A similar plan of experimentation was devised for orchard or grove 
fertilizer te<is. lilocks of four or nine trees are selected as a unit for 
each treatment and careful notes made of their condition. In the 
case of an orange grove 1 pound of the fertilizer constituents was 
applied per tree. 

In 1917 the Maine Agricultural Experiment Station at Aroostook 
Farm, near Preique Isle, Maine, inaugurated a fertilizer experiment, 
Using our triangle system and employing 20 percent differences in the 
fertilizer combinations, as shown in figure 29. The fertilizer carriers 
in die regular 21 combinations were the phosphate as acid phosphate, 
the potash ai potassium Sulfate, and the nitrogen as one-third sodium 
nitrate and two thirds ammonium sulfate. In addition a certain mix- 
ture wa ebosrn and changes made to include other carriers, as fol- 
!"••. : 1 r/j I ; or the acid phosphate an equivalent amount of floats is 
-ubstituted, and ( l> ) for the inorganic nitrogen an equivalent amount 



Journal of the American Society of Agronomy. 



Plate 6. 




Fig. 2. General view of the triangle fertilizer experiments on the Arlington 
Farm, Virginia. Harvesting the wheat. 



Journal of the American Society of Agronomy. Plate 7. 




FlG. i. General view of triangle fertilizer experiments on potatoes at Presque 
Isle, Maine. The treatments were arranged in double rows. 




I k j. View of triangle fertilizer experiments on potatoes al I'rcsquc Isle, 
Maine. Note the plants dyitiK in the two rows on the left, one of the treat- 
ments without potash. Note ;ils<. the scries of darker rows in pairs occurring 
at intervals. These are the other no-potash treatments. 



SCHREINER & SKINNER: FERTILIZER EXPERIMENTS. 245 

of organic nitrogen is substituted, both as dried blood and as high- 
grade tankage. The crops grown are potatoes, oats, and clover in 
rotation, each crop being grown each year on one of the series of 
plats. The plats are one-fortieth acre in area. The total quantity of 
active fertilizer constituents, P 2 5 , NH 3 , and K 2 applied per acre is 
240 pounds to potatoes, 80 pounds to oats, and 40 pounds to clover 
The general scheme of the field arrangement of the plats is shown in 
figure 41. 

A further study in which the triangle system was used was in con- 
nection with the potash hunger of the potato. In this experiment the 
21 treatments were in two rows each of such length as to make one- 



" O 'co-o ' 








CAocJ 


" t O- -.o " 


0-*o- to 


// S 

Z0-60- 20 


"o-*o-io ' r 


V ti 


22 

CAeeA 






lf 20 i. u * 


*0 +0 30 ' 




it u 


" J- 1.- T 11 


" (t-20-20 


S0-0-2O 


33 20-H> *o ' 






Ji " 






* „.,.-„ • 






tO 20 O 






CAocA 








» ~-o-'o " 






s- » I 


10-20-0 


CAoci 




st ^ 






u ■> 
20-0-to 




I* 3 








Ti ■ X 




ir z — ', — 37 






20 20-U 


W 71 






71 IS 

TJX"-" 








" (0 « " 


e»«.t 












' ttoU 


'* O- 10-20 






4s to " 










4*-c-it 


'•» 3<> „ ' 










'" , o 






















1*a " 


































/jroostiQk /arm / 9/7 




7TT ~ - Ic 






Aaint 


Experiment /Station 



Fig. 41. Plat of the triangle fertilizer experiment at Aroostook Farm of the 
Maine Agricultural Experiment Station. Plats not arranged in triangle form. 



twentieth of an acre for each treatment. Every three treatments or 
six rows were separated by the insertion of two check rows. The ex- 
periment was laid out in a field of which a carefully surveyed and 
detailed soil map had been prepared. Two soil types with gradations 
of each were involved and the crop was in each case on both soil 
types. Each soil type was harvested separately for each treatment. 
The details of this work and the highly interesting and different re- 
sults obtained for the two soil types will be reported upon later. 15 
Plate 7, figure I, shows this field triangle experiment at blooming 
time. The characteristic symptoms of potash hunger are a deeper 
green foliage of the plants than is normal, with a crinkled appearance 
of the leaf, which curves downward. This darker green gives way 
later to a distinct bronzing of the leaves, which finally die and the 

15 This was a cooperative study of potash hunger of the potato with special 
reference to soil type differences, carried on by the Bureau of Plant Industry 
and the Maine Agricultural Experiment Station. 



246 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



plant collapses completely. Plate 7, figure 2, shows some of the col- 
lapsed plants on one of the no-potash plats of the triangle (No. 6) in 
the foreground. Attention is called to the series of darker lines of 
two-row treatments which can be distinctly seen even in the photo- 
graph and in the field these were very striking indeed. These are 
the no-potash plats of the triangle with the darker green or bronze- 
colored foliage characteristic of this trouble. This triangle exper- 
iment brought out strikingly the effect which. potash in the fertilizers 
has in controlling this war-time disease, and shows how it is practical 
to use such a system in farm experimentation for solving fertilizer 
needs for any particular soil or crop. 



RELATIVE EFFECT OF SODIUM CHLORIDE ON THE 
DEVELOPMENT OF CERTAIN LEGUMES. 1 

G. W.. Hendry. 
Introduction. 

The experiment here reported was initiated to observe the relative 
reaction to NaCl of certain leguminous crops cultivated extensively 
in California. Because of the complexity of the problem no effort 
has been made to arrive at the absolute tolerance of any variety and 
even the relative tolerance, indicated under the conditions of this ex- 
periment/ may not be maintained under changed conditions or with 
other salts or combinations of salts. 

Thirteen selected varieties were grown in the greenhouse in i-quart 
glass jars filled with chemically pure quartz sand, to which alkali in 
the form of NaCl was added in five different concentrations, viz., 
2,000, 8,000, 15,000, 25,000. and 50,000 parts per million of solution, 
equivalent to 0.04 percent, 0.16 percent, 0.3 percent, 0.5 percent, and 
1.0 percent respectively of alkali based on the dry weight of the sand. 
In addition each jar was supplied from time to time with plant food 
in the form of a standard nutrient solution. Each variety was also 
grown in a culture receiving no NaCl. Sprouted seeds of all varieties 
Were placed in the Band Oil September 10, but the seeds were placed 
in the terminator in preparation for the experiment as follows: Wind- 
sor bean and garban/.o, September 3; Cranberry, Red Kidney, Bayo, 

1 Contribution from the hivision of Agronomy, College of Agriculture, Uni- 

rei • • oi 1 iliforaJ*, Berkeley, Cal. Received for publication April 14, [9x8. 



HENDRY : EFFECT OF SALT ON LEGUME GROWTH. 



247 



and lima beans, September 4; Pink, Red Mexican, Lady Washington, 
Small White, and Blue Pod beans, September 5 ; and Blackeye cow- 
pea and Tepary bean, September 6. By this progressive arrangement 
the radicles of all varieties were about 1 inch in length when planted 
September 10. 

Relative Effects of NaCl Concentration on Plant Development. 

In Table 1, the varieties are grouped in the order of their apparent 
resistance to NaCl, and observations are recorded on the comparative 
effect of alkali concentration on the duration of life, height growth, 
and total leaf area. 



Table i. — Relative effects of NaCl concentration on life period, height growth, 
and leaf area of certain leguminous crops. a 





No. of davs clams 


ived 


Height in inches when 


Relative leaf surface, 6 














32 days old. 






percent. 




Class and variety. 




NaCl 


NaCl 


NaCl 




NaCl 


NaCl 


NaCl 




NaCl 


NaCl 


NaCl 


A 


0.04 


0.16 


o-3 




0.04 


0.16 


o-3 


J, 


0.04 


0.16 


°-3 






per- 


per- 


per- 


J: 


per- 


per- 


per- 




per- 


per- 


per- 




U 


cent. 


cent. 


cent. 


O 


cent. 


cent. 


cent. 


O 


cent. 


cent. 


cent. 


Most tolerant: 


























Blackeye cow- 


























pea 


C Q0 


c QO 


c go 


c go 


14.O 


14.O 


6.5 


4-5 


100 


94.O 


16.7 


8.0 


Windsor bean . . 


c go 


c go 


c go 


67 


32.0 


22.0 


19.0 


9.0 


100 


Co.o 


30.7 


9-3 


Mexican gar- 


























banzo 


c go 


c go 


66 


59 


23.0 


25.O 


l6.0 


II.O 


100 


76.4 


50.0 


15-0 


Moderately toler- 


























ant: 


























Lewis lima 


c QO 


c go 


c go 


50 


54-0 


27.0 


IO. O 


6-5 


100 


5i-o 


20.0 


3-1 


White tepary. . . 


d go 


d go 


d go 


44 


52.0 


42.0 


32.0 


6.7 


100 


75.2 


27-5 


6.0 


Least tolerant: 


























Cranberry 


rf 8 5 


d 85 




40 


42.0 


32.0 




6.0 


100 


5i-9 






Small white. ... 


d go 


d go 


d go 


3i 


52.0 


32.0 


7.0 




100 


98.0 


29-5 




Red kidney .... 


d go 


d go 


7i 


31 


19.0 


20.0 


10.5 




100 


40.5 


12.7 




Lady Washing- 




























d 86 


d S6 


56 


3i 


27.0 


II. 


6-5 




100 


61. 1 


14-7 




Pink 


d 76 


d 8i 


50 


3i 


11. 


9-5 


7-0 




100 


40.0 


8.4 




Red Mexican . . 


rf 86 


d 86 


50 


3i 


10. 


9-5 


5-0 




100 


71.0 


4-7 




Bayo 


d go 




61 


15 


29.0 




2-3 




100 




47-1 




Blue pod 


d go 


d go 


46 


10 


48.0 


36.0 


6.0 




100 


90.5 


17.8 





None of the plants grew in the 1.0 percent solution of NaCl. In the 0.5 
percent solution only the Windsor bean and the Mexican garbanzo survived, 
both of them living for 45 days. The former reached a height of 1 inch and 
the latter of 4 inches, while the relative leaf surfaces were reduced to 3.6 and 
2.0 percent of their respective checks. 

6 The relative leaf surface is expressed in terms of leaf coefficient or average 
number X average length X average width. Taken when plants were 46 days old. 

c Still growing when experiment was stopped. 

d Ripened seed. 



24^ JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



The grouping's employed in Table I are based principally upon the 
data contained therein, but were influenced somewhat, particularly in 
the arrangement of the varieties within the groups, by the more in- 
tangible qualities of general thrift and vigor. It is significant that 
each of the varieties in the first (most tolerant) group, Blackeye cow- 
pea (Vigna sinensis), Windsor bean (Vieia faba), and Mexican 
garbanzo (Cicer arietinnm), represents a different genus from any 
of the others ; and that the varieties in the second group (moderately 
tolerant), Lewis lima (Phaseohts lunatus) and White tepary (P. 
acutifolius var. latifolhis Freeman), both represent different species 
from those in the last group (least tolerant), all of which are varieties 
of Phase olus vulgaris. 

Relative Effects of NaCl Concentration on Nodule Development. 

Xodules of nitrogen-fixing bacteria developed naturally on all 
varieties excepting the lima, tepary, and garbanzo, and were most 
numerous and largest in each instance in the check culture, diminish- 
ing in number and size as the strength of the solution increased, and 
disappearing entirely in the soils containing NaCl equivalent to 0.3 
percent and 0.5 percent on the dry basis. The fact that the nodules 
on the Windsor bean were apparently less injured by NaCl than those 
on the Blackeye cowpea suggests that the organisms themselves may 
possess specific alkali tolerance. Some observations bearing upon this 
relationship are given in Table 2. 



TABLE 2. — Relative effects of NaCl concentration on nodule development. 



Variety. 


Average number of nodules 
per plant. 


Average diameter of nodules 
in mm. 


Check. 


NaCl 
0.04 per- 
cent. 


NaCl 
o.x6 per- 
cent. 


NaCl 
0.3 per- 
cent. 


Check. 


NaCl 
0.04 per- 
cent. 


NaCl 
0.16 per- 
cent. 


NaCl 
0.3 per- 
cent. 




110 

$0 

40 
10 


85 
35 


50 
6 

15 



c c c 


2.0 

3-75 

5-5 

3-8 


4.0 
2.0 


2.1 

1.5 
3-0 























1 









RlLATXVl l.i 1 1< is of NaCl Concentration on Blossoming PkRien. 

'! he occurrence of the blossoming period was retarded by the 
I rescuer r.f Nad in the soil, and the period of retardation increased 
Bi the concentration Of -.'tit became greater. The retardation was 
more in some varieties than in others. The Pink bean, however, 
was an exception, blossoming simultaneously in all cultures, includ- 



HENDRY: EFFECT OF SALT ON LEGUME GROWTH. 249 

ing the check, independently of salt concentration. There was no 
apparent correlation between the period of retardation in blossoming 
and alkali resistance. Observations on the blossoming periods of six 
varieties are recorded in Table 3. 



Table 3. — Relative effects of NaCl concentration on blossoming periods of 

several legumes. 



Variety. 


Date of opening of the first 

Check NaC1 

0.04 percent. 


blossoms. 

NaCl 
0.16 percent. 


Total period of 
retardation in 
0.16 percent 
solution. 








Days. 


Windsor bean 


Oct. 12 Nov. 5 


Nov. 8 


27 


White tepary 


Oct. 20 Oct. 22 


Oct. 26 


6 


Red kidney 


Oct. 12 Oct. 30 


Nov. 3 


22 


Pink 


Oct. 8 Oct. 8 


Oct. 8 





Lady Washington 


Oct. 17 Oct. 27 


Nov. 10 


24 


Small white 


Oct. 22 Oct. 25 


Nov. 8 


17 



Summary. 

The Windsor bean (Vicia faba), the Blackeye cowpea (Vigna 
sinensis), and the Mexican garbanzo (Cicer arietinum), were less 
affected by NaCl than the other varieties tested. 

The Lewis lima (Phaseolus lunatus), and the White tepary (P. 
acutifolins var. latifolius Freeman), were less affected by NaCl than 
any of the varieties of Phaseolus vulgaris tested. 

The visible effects of NaCl upon the development of the plants was 
as follows : 

1. Retardation of germination; 

2. Retardation of height growth ; 

3. Reduction of the number of leaves ; 

4. Reduction of the size of the leaves ; 

5. Retardation of the blossoming period; 

6. Reduction in the number of nodules; 

7. Reduction in the size of the nodules ; and 

8. Premature death. 



250 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



RELATION BETWEEN YIELD AND EAR CHARACTERS IN 

CORN. 1 

T. B. Hutcheson and T. K. Wolfe. 2 

In recent years some doubt has arisen as to the value of certain 
score-card points as a criterion for selecting high-yielding strains of 
corn. The question is, do the points emphasized on the score card 
have any relation to the yielding capacity of the individuals possessing 
these characters? It would be of great importance to determine 
which of these points are and which are not indicative of high yield, 
if any relation at all exists. Considerable work has been done along 
this line, some of which will be reviewed briefly in this paper. 

The usual method in studying the relation of ear characters to 
yield has been to select ears with the characters desired and then ob- 
tain the yield from these selected ears when planted. In this method 
of procedure, there is likely to be some variation in the yield of dif- 
ferent ears, due to cross-pollination. It is well known that the effect 
of broad breeding in corn is often marked. It may be that this 
effect has overshadowed differences in yield due to different charac- 
ters possessed by the selected ears. 

The data in this paper deal with the relation between yield and ear 
characters of the progeny of certain seed ears selected at random. 
The characters studied in relation to yield are average length, aver- 
age circumference, ratio of tip to butt circumference, average circum- 
ference of cob, percentage of grain, average number of rows, aver- 
age length of kernels, uniformity of exhibit, shape of ears and trite- 
ness to type, character of tips, character of butts, uniformity of 
kernels, space between kernels, and space between rows. 

IvKVII.W OK LlTKRATURE. 

[n extensive experiments conducted at the Ohio (io) 3 station it 
was found that there was no material relationship between various 
(•<•<} eai liaracters and \ ield. 

1 Paper No. 4, Department of Agronomy, Virginia Agricultural Experiment 

Hi, I 1 ; 1 f k' :)>\irn, Va. Receive! for publication April 28, 1918. 
1 I he writers wish to acknowledge the valuable assistance of Mr. S. C. Ilar- 
man in obtaining the measurements reported in this paper. 

.'.'miller in parentheses refer to " Literature cited," p. 255. 



HUTCHESON & WOLFE : CORRELATIONS IN CORN. 



2 5 I 



Love (4) obtained a slight increase in yield from planting long 
ears and from planting heavy ears. However, such seed ear charac- 
ters as number of rows, average weight of kernel, and ratio of tip to 
butt did not have any marked effect on yield. 

Love and Wentz (5) studied the relation of such seed ear charac- 
ters as length, average circumference, average cob circumference, 
weight of ear, number of rows, average weight, average length, aver- 
age width of kernels, and percentage of grain to yield. The average 
circumference of the seed ear was the only character which showed 
any significant relation to yield. The writers conclude that " the 
only basis left for selecting high-yielding seed corn is the ear-to-row 
progeny test." 

Hartley (2), studying four varieties of corn over a period of six 
years, in which more than 1,000 ear-to-row tests of production were 
made, obtained results indicating that no visible characters of appar- 
ently good seed ears are indicative of high yielding capacity. 

Pearl and Surface (8), in a two years' ear-to-row test, found that 
there was no evidence of any close association or correlation between 
the size and conformation of the seed ear and the yield of corn ob- 
tained from it on planting. 

McCall and Wheeler's (6) experiments indicate that neither length, 
weight, nor density of ear is correlated with yield. 

Sconce (9), in studying the relation between various seed ear char- 
acters to yield in the Reid Yellow Dent and Johnson County White 
varieties, found that ears containing 18 or 20 rows gave the highest 
yield. In Reid Yellow Dent, small-germ kernels gave the best re- 
sults, but the large-germ kernels of Johnson County White gave the 
highest yield. The relation of shape of kernel and yield is striking 
in both varieties used. The writer states : " The kernel of ideal shape, 
which tapers slightly and has the square shoulders and full tip, has 
been giving the best results. Not once since beginning the experi- 
ment has an ill-shaped kernel on the average outyielded the ideally 
shaped kernel." 

Montgomery (7) found that the long, smooth type of seed ears 
outyielded the standard type ears. Also, extra large ears are no more 
valuable than medium-sized ears for seed purposes. 

In an experiment conducted at the Iowa station (3) 500 ears of 
corn were secured from the field without any selection and scored by 
twenty-five judges. A portion of each ear was planted in the field; 
the first year's results indicate that the ears receiving the highest 
scores were also the most productive in the field. As compared with 



252 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



the bulk of the ears, the fifty best ears, as selected by the majority 
of the judges, yielded on the average 5 bushels more to the acre. 

Experiments conducted by Cunningham ( 1 ) indicate that the length 
of ear has little relation to yield, but that varieties differ in this re- 
spect. The indications are that slender seed ears are more produc- 
tive than those of comparatively large circumference. There was ap- 
parently no relation between the character of tips and butts and per- 
centage of grain to cob to yield. It was found that ears of inter- 
mediate indentation outyielded smooth or rough ears, while the rough 
consistently yielded lower than the smooth ears. The relation of 
number of rows to yield varied with different varieties. 

Material and Methods. 

The corn used in this work was Boone County White, which has 
been grown at the experiment station for nine years and selected for 
earliness. By selection a strain has been developed adapted to the 
mountain sections of the State. This strain matures about ten days 
to two weeks earlier than other strains of this variety which have 
not been selected for earlines. 

All measurements were taken in inches and the yield expressed in 
bushels per acre. In 1916, 140 ears were planted in the ear-to-row 
test; in 1917, 98 ears were planted. These ears were selected at ran- 
dom and two rows 66 feet long were planted from each ear, the two 
rows constituting one one-hundredth of an acre. At husking time the 
grain from the two rows produced by each ear was combined and later 
shelled and the yield per acre computed. Each year, before shelling, a 
certain number of high-yielding and low-yielding strains were selected 
and the data secured which are presented in this paper. The average 
circumferences of the ear and of the cob were obtained by averaging 
the butt and tip circumferences. The ratio of tip circumference to 
butt circumference was obtained by dividing the former by the latter. 
In this way the shape of the ear was determined. The nearer this 
ratio approaches unity the more cylindrical is the ear. The percen- 
tal- of grain was calculated by dividing the weight of shelled corn 

by the weight of grain and cob. The number of rows was deter- 
mined by counting the rows on ten ears of each strain selected. The 
ITerage length of Kernels was found by subtracting the average cir- 
cumferenee of the cob from the average circumference of the ear. 
The length of the kernels was then calculated by the formula, cir- 
1 Ufl ) f erence- \2wf (f radius OT length of kernel). Each strain 
• <■]<< \r<\ u;i<, scored by use of the corn score card adopted by the 



HUTCHESON & WOLFE : CORRELATIONS IN CORN. 



253 



station and corn growers of this State. Data were secured by use of 
the score card on the following characters : Uniformity of exhibit, 
shape of ears and trueness to type, character of tips, character of 
butts, uniformity of kernels, shape of kernels and size of germs, 
space between rows, and space between kernels. The value of these 
latter characters is expressed in percentages, 100 percent being a per- 
fect score. 

Results. 

In Table 1 data are presented showing the relation between yield 
and various ear characters of the progeny of a number of ears of 
Boone County White corn planted in the ear-to-row test. 



Table i. — Relation between yield and various ear characters of the progeny of 
different ears of Boone County White corn at the Virginia station 
in 19 16 and 19 17. 



Character. 


High yielding strains. 


Low yielding strains. 


Average. 


1916. 


1917. 


1916. 


1917. 


High 
yielders. 


Low 
yielders. 


Average length, in 


8.6l 


8.31 


8.30 


7.65 


8.46 


7.98 


Average circumference, in. . . 


6.9O 


6.70 


6.69 


6.46 


6.80 


6-57 


Ratio of tip to butt circum- 














ference 


•85 


.88 


.84 


.87 


.86 


.86 


Average circumference of 














cob, in 


4.09 


4.29 


3-97 


4.08 


4.19 


4-03 


Yield, bushels per acre 


82.74 


62.30 


61.97 


47-58 


72.52 


54-78 




85.27 


79.46 


83-97 


' 80.57 


82.37 


82.27 


Average number of rows .... 


16.77 


16.51 


16.57 


16.35 


16.64 


16.46 


Average length of kernels, in. 


•45 


•38 


•43 


•38 


.42 


.41 


Uniformity of exhibit, per- 
















55-oo 


49.44 


48.00 


48.18 


52.22 


48.09 


Shape of ear and trueness of 














type, percent 


55-42 


41.67 


50.50 


38.64 


48.55 


44-57 


Character of tips, percent. . . 


60.83 


26.11 


47.00 


25-50 


43-47 


36.25 


Character of butts, percent. . 


59-17 


39-44 


55-00 


40.91 


49-3 1 


47.96 


Uniformity of kernels, per- 
















55-83 


43-89 


56.00 


35-45 


49.86 


45-73 


Shape of kernels and size of 














germ, percent 


64-58 


40.56 


53.50 


40-45 


52.57 


46.98 


Space between kernels, per- 
















71.25 


53-33 


71-50 


52.27 


62.29 


61.89 


Space between rows, percent. 


69-58 


45-56 


65.00 


52.73 


57-57 


58.87 



In 19 1 6 twelve high-yielding and ten low-yielding strains were se- 
lected. The high-yielding strains contained 756 ears which were 
measured for length and 716 ears measured for circumference. The 
low-yielding strain contained 439 ears which were measured for. 
length and 418 measured for circumference. 

In 19 1 7 nine high-yielding and eleven low-yielding strains were 



2 54 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



selected. In the high-yielding strains 514 ears were measured for 
length and 500 ears for circumference. In the low-yielding strains 
540 ears were measured for length and 534 for circumference. 

A study of Table 1 shows that in 1916 the high-yielding strains 
produced 20.77 bushels more per acre than the low-yielding strains. 
Also, the data obtained in that year are in favor of the high-yielding 
strains in every case save two, uniformity of kernels and space be- 
tween kernels. In these two instances, the differences are slight, 
0.17 percent, and 0.25 percent respectively. In 1917, the difference 
in yield is 14. 72 bushels. In this year the results are in favor of the 
high yielding strains in all instances save three, namely, percentage 
of grain, character of butts, and space between rows. The differ- 
ences in these three instances are 1.11 percent, 1.47 percent and 7.17 
percent, respectively. If the averages for the two years are consid- 
ered, we find that the results are in favor of the high-yielding strains 
in every instance, except in the case of the space between rows. 
The difference here is 1.30 percent in favor of the low-yielding 
strains. 

The negative relation between yield and space between rows is not 
surprising when the origin of the strain of corn is considered. As 
stated before, the strain has been developed especially for early ma- 
turity and in so doing a type has been secured with a rather greater 
distance between the rows than in the original variety. 

The results for the two years are in close accord and are very sug- 
gestive. The data indicate that the points emphasized in the corn 
score card may be of value in selecting high-yielding strains. In 
other words, according to our results, high-yielding strains are high- 
scoring strains. 

These results are more interesting when the prize-winning varieties 
of corn in Virginia arc considered. For many years the standard 
varieties of com in the State, based on yield and other desirable 
characteristics, have been Moone County White, Johnson County 
White, Collier Excelsior, Reid Yellow Dent, and Gold Standard. 
These varieties arc not only the high-yielding varieties of the State, 
but arc the ones which have taken the greater part of the prizes an- 
nually at the State fairs and at the fairs held by the State Corn 
Growers' Association, where the score card is used as a basis for 
awards. 

Conclusions. 
The data reported in this paper indicate that: 

1. The relation between yield and length, average circumference, 



HUTCHESON & WOLFE I CORRELATIONS IN CORN. 



255 



average circumference of cob, uniformity of exhibit, shape of ears 
and trueness to type, character of tips, uniformity of kernels, and 
shape of kernels and size of germ is significant. 

2. The relation between yield and ratio of butt to tip circumfer- 
ence, percentage of grain, number of rows, average length of kernels, 
character of butts, space between kernels, and space between rows is 
small. 

3. The points emphasized on the score card are of value in select- 
ing high-yielding strains of corn. 

4. High-yielding strains of corn are high-scoring strains. 

Literature Cited. 

1. Cunningham, C. C. 

The relation of ear characters of corn to yield. In Jour. Amer. Soc. 
Agron., v. 8, no. 3, p. 188-196. 1916. 

2. Hartley, C. P. 

Progress in methods of producing higher yielding strains of corn. In 
U. S. Dept. Agr. Yearbook for 1909, p. 309-320. 1910. 

3. Hughes, H. D. 

An interesting seed corn experiment. In Iowa Agr., v. 17, no. 9, p. 424, 
425, 448. 1917. Abs. in Expt. Sta. Rec, 37: 830. 1918. 

4. Love, H. H. 

The relation of certain ear characters to yield in corn. In Proc. Amer. 
Breeders' Asso., 7 : 29-40. 1912. 

5. Love, H. H., and Wentz, J. B. 

Correlations between ear characters and yield in corn. In Jour. Amer. 
Soc. Agron., v. 9, no. 7, p. 315-322. 1917. 

6. McCall, "A. G., and Wheeler, Clark S. 

Ear characters not correlated with yield in corn. In Jour. Amer. Soc. 
Agron., v. 5, no. 2, p. 117-118. 1913. 

7. Montgomery, E. G. 

Experiments with corn. Nebr. Agr. Expt. Sta. Bui. 112. 1909. 

8. Pearl, Raymond, and Surface, Frank M. 

Experiments in breeding sweet corn. In Ann. Rept. Maine Agr. Expt. 
Sta., p. 249-307. 1910. 

9. Sconce, H. J. 

Scientific corn breeding. In Proc. Amer. Breeders' Asso., 7 : 43-50. 1911. 
10. Williams, C. G., and Welton, F. A. 

Corn experiments. Ohio Agr. Expt. Sta. Bui. 282. 1915. 



256 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



INFLUENCE OF CEROTOMA TRIFURCATA ON THE NITROGEN- 
GATHERING FUNCTIONS OF THE COWPEA. 1 

Lewis T. Leonard and C. F. Turner. 
Introduction. 

The nodules of leguminous plants are recognized as fundamental 
sources of nitrogen for plant growth and any agency which tends to 
impede their normal formation or interfere with their nitrogen-fixing 
functions is worthy of scientific consideration. For a long time it has 
been known that the physical, chemical, and bacteriological forces of 
the soil influence the production of nodules, but recently McCon- 
nell 2 called attention to the destruction of the nodules of various 
legumes by the larvae of the bean-leaf beetle, Cerotoma trifurcata 
Forst. McConnell has also reported 3 the destruction of nodules of 
wild legumes by the larvae of Eudiagogus rosenschocldi Fahrs in 
Mississippi and Arkansas. The first-mentioned beetle in its aduft 
form is very destructive to the leaves of the legumes which it will 
attack, but this injury is quite noticeable, whereas the injury done 
by the larvae to the roots and tubercles represents a type which is 
not evidenced by any superficial symptoms which the ordinary ob- 
server might consider (Plate 8, fig. 1). The cowpea and the garden 
bean are reported to be the principal plants which are subject to the 
ravages of these insects. As the cowpea is one of the most important 
legume crops in the sections where the beetles arc apparently most 
destructive it was chosen for use in these experiments to determine 
the factors governing this type of injury. 

It wa- at first thought that results could be obtained by conducting 
experiments in the vicinity of Washington D. C, and in 19T4 field 
and can experiment^ were carried on at the Arlington Farm, Rosslyn, 
Ya., but the results were vitiated by a cold, rainy season. Similar 
experiment! of a more comprehensive character were conducted in 

1 ContrilMitK.ii from the Bureaus of Plant Industry and Entomology, U. S. 

Department of Agriculture, Washington, D, C. Received for publication April 
5. 1918. 

' • '••••*; \V. K A unique type of inseel injury. In Jour. Econ. Knt., 

8: 2f>i 267. [pi 5. 

\M onnell, W. R. Another nodule destroying beetle. In Jour. Econ. Ent., 

8: 551. ICM5. 



LEONARD & TURNER: BEETLE INJURIES TO COWPEA NODULES. 257 

1915 at the same place, but the results did not justify their continua- 
tion. It was evident that climatic conditions obtaining in the latitude 
of Washington were not conducive to the normal life of the insect, 
regardless of the fact that Cerotoma bettles have been collected in 
this vicinity^ and as far north as New Jersey. The beetles used in 
these initial trials were collected in Mississippi and the eggs were ovi- 
posited at the Bureau of Entomology Field Station, Hagerstown, Md. 
It is possible that the strains of larvae and beetles developing under 
abnormal conditions lacked the virility to function properly in the 
northern climate. 

As the results at Washington were inconclusive and entirely un- 
satisfactory, it was decided to continue the experiments in the vicin- 
ity of Greenwood, Miss., under the direct supervision of the Federal 
entomological substation there. 

Experimental Data. 

These experiments were carried out during the summer of 1916. 
Similar plots were arranged at Greenwood, Miss., on the rich black 
delta soil and at Grenada, Miss., on the poorer reddish hill soil. 
The location of these plots was determined on the basis of soil uni- 
formity, contour of land, and proximity to fields planted or to be 
planted in cowpeas. These plots were cleared of all debris consider- 
ably before planting time so as to make the hibernating insects seek 
other quarters or die from exposure. Each plot was plowed and 
subdivided into smaller plots 4 feet square centered in 12- foot squares, 
thereby leaving a clear space 4 feet in width around each small plot. 
Before breaking ground the Greenwood plot was in grass and weeds 
while the one at Grenada was in clover. 

Cages to protect the small plots from outside infestation and to 
restrict the travel of beetles introduced or developed in the cages 
were made from ^4 -inch lumber and 18-mesh wire screen. The out- 
side dimensions of the cages were 4 feet by 4 feet by 2^ feet. They 
were strongly built and well braced. 

These cages were arranged in the order shown below. 



Final Arrangement of Experimental Plots at Greenwood, Miss. 



1 


2 


3 


4 


5 


6 


7 


A 


B 


C 


O 


A 


B 


C 


14 


13 


12 


11 


10 


9 


8 


A 


C 


A 





A 


C 


A 


15 


16 


17 


18 


. 19 


20 


21 


C 


A 


B 





C 


A 


B 



258 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Final Arrangement of Experiment Plots at Grenada, Miss. 



I 


2 


3 


4 


5 


6 


7 


A 


c 


B 





A 


B 


C 


14 


13 


12 


11 


10 


9 


8 


A, 


c 


A 


O 


A 


C 


A 


15 


16 


17 


18 


19 


20 


21 


C 


A 


B 





C 


A 


B 



The smaller plots were numbered for convenience in handling 
samples. 

Plots marked A were entirely inclosed in cages and twenty pairs of 
Cerotoma beetles were introduced into each except those in the center 
row bearing the numbers 8, 10, 12, and 14, into each of which forty 
Cerotoma eggs were inserted. 

Plots marked B were covered with cages which were slightly ele- 
vated above the ground on one side to allow for normal infestation 
in conjunction with plants shaded by cages (Plate 8, fig. 2). 

Plots marked C were inclosed the same as those marked A, but 
were not infested with beetles or eggs. These plots were devised to 
act as checks on the inclosed infested plots. 

Plots marked O were not covered and served the purpose of open 
controls. 

The cages from which it was intended to prevent outside infesta- 
tion by beetles or larvae were sunk 6 inches in the ground up to the 
top of the baseboard and precautions were taken to make them secure 
by filling all cracks and holes. 

New Era cowpeas were planted in the small plots at Greenwood 
May 6, 1916, and at Grenada, May 25, 1916. Infestation was ac- 
complished in accordance with the plans at Greenwood, June 14, 1916, 
and at Grenada, June [6, 1916. At the time of infestation the num- 
ber of plants per plot was reduced to six. The crop was harvested 
at Greenwood, July 22, \<)\C), and at Grenada, August 2, 1916. The 
cowpea vines were cut about 3 inches from the ground. This forage 
was stored in cotton bags and dried in the sun. At the same time 
the crop was cut the roots were dug carefully and data taken on the 
total number of nodukfl per plot and the percentage of nodules in- 
jured by Cerotoma larvae. The roots were also dried. 

In the second series of experiments the plots at Greenwood were 
planted with Whippoorwill cow peas August 7, 1916, and those at 
Grenada, AugUfl <>. I'yio. [nfestatidfl in exactly the same quantities 
and mann« r as in the earlier experiments was introduced at Green- 



THE LIBRARY 
OF THE 



Journal of the American Society of Agronomy. 



Plate 8 




Pm I |rpc of ; < n ed $Xl6 method of tilting to secure natural infestation. 



LEONARD & TURNER: BEETLE INJURIES TO COWPEA NODULES. 2$C 



wood, September 2, 1916, and at Grenada, September 5, 1916. As in 
the other experiments the number of plants were reduced to six. 

These experiments were concluded at Greenwood October 17, 1916, 
and at Grenada October 18, 1916. Except for the collection of com- 
posite soil samples for nitrogen and bacteriological analyses the crop 
was treated the same as before and similar data obtained. 

The material was dried in Mississippi and shipped to Washington, 
where it was carefully weighed and ground to a powder. The 
Bureau of Chemistry kindly made the nitrogen analyses recorded in 
Table 1. 



Table i. — Summary of results obtained in experiments to determine Ceretoma 
injury to cowpeas in 1916. 

New Era Cowpeas at Greenwood, Miss. 



Plot.i 


Soil ni- 
trogen, 
percent. 


Soil-nitrify- 
ing bacteria 
per gram. 


Ave. 
number 

of 
nodules 
per plot. 


Percent Ave 

°} . total dry 
nodule we ight 
injury. 


Per 

Of Iiitl 

Tops. 


cei.t 
ogen. 

Roots. 


Total 
nitro- 
gen, 
grams. 


A (Introduced beetles) . 
B (Normal infestation). 

C (Covered check) 

A (Introduced eggs) . . . 
O (Open check) 


O.17 
0.18 
O.24 
O.38 
0.16 


11,687,500 
14,250,000 
7,000,000 
5,750,000 
12,500,000 


207.50 
263.50 
373-33 
284.50 
442.67 


95-35 278.38 

67.08 240.38 
42.37 238.83 
59.63 197-38 
93.40 457.17 


2.40 
2.80 
2.67 

2-75 
2.60 


1.28 

1-55 
I.72 
I.72 
I.5I 


6.32 
6.56 
6.24 
5.26 
II.64 



New Era Cowpeas at Grenada, Miss. 



A (Introduced beetles) . 
B (Normal infestation). 

C (Covered check) 

A (Introduced eggs) . . . 
O (Open check) 


0.04 
0.03 
0.06 
0.03 
0.04 


245,000 
375,000 
215,000 
232,500 
370,000 


217.00 
183.25 
173.67 
177.50 
198.00 


76.18 
61.10 
60.59 
65-53 
79 03 


73-88 
110.25 
97-17 
93-13 
85.80 


2. 11 1.08 
2.42 1. 31 
2.10 1. 16 
2.47 1.26 
2.13 1.27 


I.42 
2.50 
I.90 
2.l6 
1-73 


Whippoorwill Cowpeas at Greenwood, Miss. 


A (Introduced beetles) . 
B (Normal infestation). 

C (Covered check) 

A (Introduced eggs) . . . 
O (Open check) 


0.17 
0.18 
0.24 
0.38 
0.16 


11,687,500 
14,250,000 
7,000,000 
5,750,000 
12,500,000 


102.50 
95-50 
195-17 
122.75 
6.67 


49.08 
44.88 
10.77 
21.12 
80.50 


96.38 
26.63 
121.42 
114.50 
1.67 


2- 77 

3- 07 
2-73 
2.91 


i-35 

x v 

1.60 

LSI 
0.51 


2.49 
•74 
3-i6 
3-14 
0.01 


Whippoorwill Cowpeas at Grenada, Miss. 


A (Introduced beetles) . 
B (Normal infestation). 

A (Introduced eggs) . . . 
O (Open check) 


0.04 
0.03 
0.06 
0.03 
0.04 


245,000 
375.ooo 
215,000 
232,000 
370,000 


137-75 
110.75 
150.67 
94-50 
29.67 


46.10 
20.25 
10.02 
42.88 
73.00 


22.00 
11.38 
21.63 
14.50 
3.67 


2.71 

3-II 
2.78 
2.65 
0.69 


1. 14 
I.46 
1-23 
1. 17 
O.46 


0.52 
0.30 
0-54 
o.35 
0.02 



1 In the above summary data following A, B, C, A eggs, and O represent the 
average of 4, 4, 6, 4 and 3 plots respectively. 



260 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Discussion of Results. 

As will be noted in the summary of results, the second experiment, 
in which Whippoorwill cowpeas were used, is more striking than the 
first experiment, in which New Era cowpeas were used. This is 
probably due to factors such as planting time, climatic conditions, etc., 
rather than to the variety of cowpea. 

The average of the dry weights of the New Era cowpeas from the 
open check plots at Greenwood is greater than the average weight of 
any of the other series of plots. It is evident that the plants in this 
particular set of open checks were well started and growing vig- 
orously before the larvae made an extensive attack on the roots. 
This factor coupled with climatic conditions and the natural richness 
of the soil will probably explain this apparently abnormal produc- 
tion of forage while the inclosed plots were somewhat hindered in 
their development by the unnatural conditions imposed upon them. 

It is to be regretted that the inclosed check plots became infested 
in spite of the precautions taken. The presence of infestation in 
these cages may be explained on the ground that small female beetles 
crawled through the meshes of the screen or through the cracks at 
the corners of the cages made by the drying influence of the hot 
weather. 

The results indicate that the nodule injury is related to the percen- 
tage of nitrogen in the roots. As the injury of the nodules increases 
the percentage of nitrogen in the roots decreases, or nitrogen lost is 
in proportion to the number of larvae present. It should be noted 
that there was considerable damage done to the leaves of the cow- 
peas, much more in the second series of experiments than the first. 
The maximum damage to leaves apparently occurred in the B plots, 
although leaves were injured in practically every plot. The open 
( lurk plots in the second scries produced very spindly plants and 
the injury observed here was probably due almost entirely to the 
attacks of the larvae on the roots. 

The destructive character of the damage which may be done by 
larv;e is shown rather forcibly in the experiments with Whip- 
DOOrwill COWpeafi ; the plants did not develop much beyond the cotyle- 
don -tage. 

Cowpea bacteria will fix atmospheric nitrogen in the nodules to 

the extern of approximately two thirds of the nitrogen obtained by 
•i ' plant < hindering the Whippoorwill cowpea experiment at 
Greenwood we find that the bacteria took from the air an average 
of about j. id grains of nitrogen in the inclosed checks or C plots, and 



LEONARD & TURNER : BEETLE INJURIES TO COWPEA NODULES. 26 I 

in the open check plots an average of about 0.006 grams. A com- 
parison of these two estimates will indicate an approximation of the 
damage it is possible for these insects and their larvae to perpetrate. 
There is no doubt that the larvae and beetles are high in nitrogen and 
to some extent excrete the nitrogen they consume in the soil sur- 
rounding the cowpeas on which they feed. Analyses of beetles indi- 
cated that in the air-dry state they contained 9.65 percent nitrogen, 
while air-dry larvae contained 8.73 percent nitrogen. However, as it 
is possible for these insects during their life cycle practically to kill 
cowpea plants it will be readily seen that they constitute a direct or 
indirect menace to the nitrogen-gathering power of the cowpea. 

The average total nitrogen fixed by the nodule bacteria in the 
Whippoorwill Greenwood-B plots or plots exposed at the bottom only 
is estimated to be about 0.50 gram, showing that the damage was 
limited to some extent by the covering. In the same series of plots 
it will foe noticed that the A plots which were treated with beetles 
averaged a gain of amospheric nitrogen of about 1.86 grams, the arti- 
ficially introduced beetles not doing as much damage as those which 
entered through normal infestation. The A plots which were treated 
with eggs took from the air about the same amount of nitrogen per 
plot as did the inclosed checks. 

The dry weight column follows practically the same trend as the 
total nitrogen column and the percentage of nitrogen contained in the 
tops is fairly constant when the open checks in the second series are 
not considered. 

Conclusions. 

1. Danger of extensive damage from Cerotoma beetles or their 
larvae in the vicinity of Washington is slight. 

2. Damage to the mutual nitrogen-fixing functions of the cowpea 
plant may be caused by these insect larvae without superficial indica- 
tion of such damage except the presence of the beetles and leaf 
injury. 

3. Time of planting and preseason conditions are important fac- 
tors in lessening the extent of damage. Planting should be done 
after the over-wintered beetles have laid their eggs ; for the latitude 
of Greenwood, Miss., May 1 to 15 is the proper time. Rotation of 

I crops, fall plowing, and clean culture will probably prove beneficial, 
but further work will demonstrate the efficacy of these methods. 

4. Damage may range from practically nothing to the entire de- 
struction of the plant. 

5. The damage to the nitrogen content of cowpea roots is roughly 
proportional to the number of larvae present. 



262 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



AGRONOMIC AFFAIRS. 

NOTICE OF ANNUAL MEETING. 

The eleventh annual meeting of the American Society of Agronomy 
will be held in Baltimore, Md., on November 11 and 12, 1918; the 
hotel at which the meeting will be held will be announced later. 
Those who expect to present papers at this meeting are urged to send 
in titles promptly to the Secretary, Lyman Carrier, Department of 
Agriculture, Washington, D. C, so that the program can be planned 
and printed some weeks in advance of the meeting. 

MEMBERSHIP CHANGES. 

The membership reported in the May issue of the Journal was 
649. Since that time 3 new members have been added, making the 
present membership 652. The names and addresses of the new mem- 
bers, with a correction of a name previously published and such 
changes of address as have been reported, follow. 

New Members. 

Eedman, Lewis W., Agr. Expt. Sta., College Park, Md. 
Harlow, H. C, Agr. College, Truro, N. S M Canada. 
WILKINSON, J. V., 624 Egan St., Shreveport, La. 

Correction of Address. 
EttTIDA, BUENAVKNTUIA, Linea 97, Antiguo Ent're 8 y 10, Vedado, Habana, Cuba. 

Changes of Address. 

Adams, E. L., Chico, Cal. 

BUL, EiEJOtY d.. 1 1 1 1 Temple Bldg., Toronto, Out., Canada. 

BsnrBMjpr; L P.. EL EL No. 3, Payette, Idaho* 

OH IFfl I IB, do* W., Normal Station, I farrisonl)tirg, Va. 
COWOBJL, H. P., Box 333, Fort Smith, Ark. 
IH vimi k, I-;. I'., Knt/town, Pa. 

I I'.xiu. Wtoh H., Plant Introduction Cardcn, Cliico, Cal. 
l \').s< n. W. L., Austin, Minn. 

Justin, M M , 442 State Capitol, Salt Lake City, Utah. 
Kennev. HaOT, Kansas State Agr. College, Manhattan, Kans. 
I ' • \ " ■ ". I^' j't. Plant P.rccdiiiK, Cornell Univ., Ithaca, N. Y. 
Martin, J. 11 , Harney Branch Station, Burns, Oreg. 
IllUm, P. W., Hartford, Kans. 



AGRONOMIC AFFAIRS. 



263 



Moore, Harvey L., Leonard Apts., Bellevue & Prospect Sts., Trenton, N. J. 
Newton, Robert, Woodstock, N. B., Canada. 
Olson, P. J., R. F. D. No. 1, Grafton, N. Dak. 
Osler, H. S., Court House, Ann Arbor, Mich. 

Russel, J. C, Dept. Chem., Nebr. Wesleyan Univ., University Place, Nebr. 

Shinn, E. H., 225 Main Street, Stillwater, Okla. 

Thompson, G. E., Experiment Station, Tucson, Ariz. 

Thompson, James, College of Pharmacy, U. of W., Seattle, Wash. 

Warburton, C. W., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Woodard, John, Mt. Morris College, Mt. Morris, 111. 



ROLL OF HONOR. 

The Society's honor roll of members who are serving their country 
in its military or naval forces is constantly growing. The list as here 
printed contains the names of 42 men, but no doubt it is quite incom- 
plete. The editor will appreciate items of interest regarding any of 
these men, as well as corrections in or additions to the list. Newis 
items regarding commissions granted, citations, or other matters are 
particularly useful. The names so far reported follow. 



Albert, A. R., 
Bliss, S. W., 
Brockson, W. I., 
Bruce, O. C, 
Brunson, A. M., 
Burnett, Grover, 
Cates, Henry R., 
Chapman, James E., 
Childs, R. R., 
Deatrick, E. P., 
de Werff, H. A., 
Dickenson, R. W., 
Downs, E. E., 
Ellison, A. D., 



Freeman, Ray, 
Gentle, G. E., 
Gilbert, M. B., 
Graham, E. E., 
Gray, Samuel D., 
Holland, B. B., 
Hudelson, R. R., 
Jensen, O. F., 
Kenworthy, Chester, 
Kime, P. H., 
Macfarlane, Wallace, 
Moomaw, Leroy, 
Newton, Robert, 
Palmer, H. Wayne, 



Piemeisel, R. L., 
Purington, James A., 

QUIGLEY, J. V., 

Ratcliffe, Geo. T., 
Raymond, L. C, 
Richards, Phil E., 

SCHNEIDERHAN, F. J., 

Schoonover, W. R., 
Scott, Herschel, 
Smith, J. B., 
Starr, S. H., 
Tabor, Paul, 
Towle, R. S., 
Westbrook, E. C. 



NOTES AND NEWS. 

Whitney J. Atcheson has been appointed assistant agronomist at 
the Maryland station. 

Ross L. Bancroft, formerly associate professor of soils in the Iowa 
college, is now in charge of soil extension work in that State. 

Percy B. Barker, recently head of the department of agronomy in 
the University of Arkansas, has been appointed assistant professor 
of agricultural education in the University of Minnesota. 

H. G. Bell, for the past several years with the Chicago office of 
the American Fertilizer Association, is now in charge of the recently 
opened Canadian office of the association, at Toronto. 



264 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



R. Page Bledsoe, formerly of the California station, is now in 
charge of forage crop work on the recently established station at 
Waterville, Wash. 

H. B. Cowgill, formerly plant breeder of the Porto Rico station, 
is now in charge of the sorghum sirup investigations of the Depart- 
ment of Agriculture at Fort Smith, Ark. 

R. H. Forbes, dean and director of the Arizona college and station, 
has been granted a year's leave of absence to assist the Societe Sul- 
tanienne d'Agriculture of Cairo, Egypt, in agricultural war service 
in the Valley of the Nile. 

E. J. Iddings, dean of the Idaho college of agriculture, has been 
made director of the Idaho station as well. 

Robert Newton, formerly field husbandman for the province of 
New Brunswick, has been in active service in the Canadian army 
since July, 191 5. He is now captain of E Battery, Canadian Anti- 
Aircraft Service, and is in France. 

P. J. Olson, assistant in plant breeding at the Minnesota station, 
has resigned to engage in farming in North Dakota. 

Everett P. Reed, assistant agronomist of the New York State sta- 
tion, has resigned to become a farm bureau agent in Ohio. 

W. J. Spillman has resigned as chief of the office of farm manage- 
ment, U. S. Department of Agriculture, a position which he has held 
for the past sixteen years, to become editor of The Farm Journal, 
Philadelphia. 

G. E. Thompson, formerly agronomist in the extension service in 
Kansas, is now agronomist of the Arizona station. 

P. F. Trowbridge of the department of agricultural chemistry of 
the Missouri station has been elected director of the North Dakota 
station and entered on his new duties September i. He succeeds L. 
Van Es, resigned to become head of the veterinary department in the 
University of Nebraska. 

R. O. Westley of the Iowa college has been elected assistant pro- 
fessor and A. M. Christcnscn of North Dakota has been made in- 
structor in farm crops at the Northwest School of Agriculture, 
Crook-ion, Minn., succeeding F. L. Kcnnard and O. M. Kiser, re- 
spectively. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. io. October-November, 191 8. No. 7-8 



INTERPRETATION OF FIELD OBSERVATIONS ON THE 
MOISTNESS OF THE SUBSOIL. 1 

F. J. Alway, G. R. McDole and R. S. Trumbull. 
Introduction. 

Soil investigators and agronomists do not appear to have recog- 
nized at any time the possible practical importance of field observa- 
tions on the moistness of the subsoil in dry-land regions as a guide 
to the more intelligent employment of various cultural operations. 
As the result of some limited field studies in Saskatchewan in 1904 
and 1905, one of us suggested that in that province, where the sum- 
mer fallow is very extensively employed, the ordinary farmer, pro- 
vided with a 6-foot auger, could form a fair estimate of the moisture 
conditions of his fields before the spring was sufficiently advanced to 
allow seeding. He would thus be in position to decide intelligentlv 
whether to sow grain upon his stubble fields or to summer fallow them, 
instead of being governed by the rule of " one year of fallow followed 
by two years of grain" (i, p. 339). 2 It was suggested that all pro- 
gressive dry-land farmers would eventually provide themselves with 
soil augers so that they might keep themselves informed as to the 
moisture conditions of their fields (2, p. 42). Later studies in west- 
ern Nebraska and in the Southwest have made it evident that field 
observations might be of at least equally great practical importance in 
these dry regions (3, p. 699). 

1 The work reported in this paper was carried out in 1907 to 1913 while the 
authors were members of the staff of the Nebraska Agricultural Experiment 
Station. Received for publication April 6, 1918. 

2 References are to "Literature cited," p. 278. 

265 



266 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

In the ten years which have elapsed since the publication of these 
views, there has been no definite recognition of the practical impor- 
tance of field observations, even by the agronomists of the many dry- 
land experimental substations in this country. The nearest approach 
to such recognition is contained in a very recent publication by Chil- 
cott and Cole on " Growing Winter Wheat on the Great Plains," in 
which it is stated that "a good guide to practice is to sow a large 
acreage when the soil at seeding time is wet to as great a depth as 3 
feet," and that " the depth to which the soil is wet can usually be told 
easily by inspection, the break between the wet and dry soil being 
very sharp " (9, p. 4). Regarding the method of inspection they make 
no suggestions. 

Our soil moisture studies in the semiarid portion of Nebraska, in 
which we made notations on the moisture condition of each foot sec- 
tion of the subsoil as the samples were being taken and later deter- 
mined both the moisture content and the hygroscopic coefficient (5), 
have provided data from which to attempt a quantitative interpreta- 
tion of field observations. The use of the soil auger and a record of 
field observations upon the occasion of visits to various districts in 
the drier portions of this country and Canada have served to show 
that the matter may be of practical importance in all these places. 
In regions with a humid climate such observations are of much less 
importance. Thus studies carried out in eastern Nebraska at the 
same time as those in western Nebraska convinced us that in the 
former such observations have only a very limited field of usefulness, 
while later studies made in Minnesota indicate that here they are of 
still less value. 

The moisture condition of a soil may conveniently be expressed by 
stating both the hygroscopic coefficient and the ratio of the water 
content to this. Thus, the expression, hyg. coef. = io.o, ratio = 1.7, 
indicates a moisture content of 17.0 percent, a wilting coefficient of 
15.0 8 (8, ]>. 65)1 5-0 percent of free water, and 2.0 percent of growth 
water. Km the ratios 1.0, 1.5, and 2.0-2.5 appear to indicate, re- 
spectively, the minimum to which crop plants can reduce the soil 
hire, ili'- point at which root penetration practically ceases (7, p. 
J/X), and the water retaining capacity of well-drained arable min- 
eral ioilfl (6, p. 69), such an expression makes all these relations ap- 
parent at a glance. The ratio when used alone indicates the relative 
njoi~tn<- while it combination with the hydroscopic coefficient ex- 
prr se- tlx- moisture condition. 

* The exact figure is 14.7. 



ALWAY, MCDOLE AND TRUMBULL: MOISTNESS OF SUBSOIL. 267 

Tools for Sampling Dry Subsoils. 

Anyone who has become familiar with the use of the ordinary soil 
auger in exploring the subsoils of humid regions must be struck by 
the limitations of this tool when he first tries to employ it in the drier 
lands. Usually, as the auger is withdrawn from the hole the soil 
slides off and the loosened material, instead of being brought to the 
surface for inspection, accumulates in the bottom of the boring. At 
shallow depths, 2 or 3 feet, it is possible by means of a quick jerk of 
the auger to throw T to the surface enough of the loosened subsoil to 
allow its examination for the purposes of the ordinary soil survey, 
and by repeating the operation the hole may from time to time be 
thus cleaned sufficiently to allow the successive levels to be inspected. 
With increasing depth this device becomes less and less applicable, 
and is quite useless long before the tenth foot has been reached. It 
is evident that satisfactory samples of the subsoil can not be taken by 
any tool that will not permit the withdrawal of the loosened portion 
without admixture with material from nearer the surface. For ex- 
ploration purposes this difficulty may be met by carrying along a can 
of water and from time to time pouring some of it down the hole, the 
loosened subsoil so moistened being removable without difficulty. 
This may aid in obtaining samples for chemical analysis, but not for 
moisture determinations. 

The soil tubes invented by King and improved by Briggs allow 
satisfactory sampling to a depth of 10 feet wherever coarser rock 
fragments are absent. Even longer tubes may be employed, but the 
extreme inconvenience of these, especially when they must be carried 
from place to place, will be evident. A serious objection to the use 
of soil tubes lies in the inability of the operator to observe the dif- 
ferences in moisture content as the samples are being taken. 

A very convenient tool for use at a distance from the laboratory 
is the sleeve auger or "auger with casing" which Tinsley (10, p. 57) 
devised soon after he began his soil studies in New Mexico. The 
essential feature of this is a metal sheath or sleeve which slides down 
over the bit; as the auger is withdrawn from the boring the loosened 
material is forced tightly into the sheath and so brought to the sur- 
face. By means of this tool samples satisfactory for both chemical 
and moisture determinations may be taken from the driest subsoil 
and from any depth. When provided with 2- or 3-foot extensions, it 
may, like the ordinary auger, be extended to any desired length, and 
also used without the sheath, if the latter is made removable. In very 



268 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



dry soils the sleeve is necessary, but in moist soils it is very incon- 
venient, while between these two conditions there is a stage of moist- 
ness in which both the open and the sheathed bit do satisfactory work, 
although more rapid headway can be made with the former. The 
complete field equipment for field observations where dry subsoils are 
involved includes three augers, each 3 feet long, two of them open 
and of different diameters, the third with a sleeve and of the same 
diameter as the smaller of the other two. In addition three handles, 
two small pipe wrenches, some files, and sufficient 3-foot extensions 
to reach any depth desired are needed. Sufficient tools to sample to a 
depth of 20 feet can be carried by hand conveniently in a 3-foot case. 

Methods of Sampling. 

The purpose of our record of field observations was threefold. In 
the first place they served to guide the one of us who was doing the 
sampling, our interest in learning the extremes of moistness found 
under any given tillage or crop condition being much greater than in 
ascertaining merely the average moisture conditions. To obtain sets 
of samples representative of the extremes we often first explored a 
field without taking samples, later deciding from the notations thus 
secured where and from what depths to collect the samples to be sent 
to the laboratory. 

In the second place, when only one of us was in the field, these 
notes served to keep the workers at the laboratory in close touch with 
the field conditions. Long in advance of the arrival of the samples 
the men at the laboratory had a far better idea as to the actual moisture 
conditions being encountered in the field than otherwise would have 
been possible until after the samples had been dried and subjected to 
the hygroscopic coefficient determination. The tediousness of the 
latter is such that under our working conditions many months often 
elapsed between the taking of the samples and the determination of 
this value. The indirect method of Briggs and Shantz (8) for ob- 
taining the hygroscopic coefficient from the moisture equivalent was 
developed by these authors only near the close of our work in Ne- 
a and even then, bad the rather elaborate and costly equipment 
■ ! been at our di-po ;il. almost as much delay would have been 
was the case with the direct method. The Simpler and 
more 1 editions indirect method, based upon a determination of the 
hygroscopic moisture ( .], p. 351 ), had not then been developed, it 
being an outgrowth of our moisture studies. 

Li tly, we had it in mind to attempt a numerical interpretation 



ALWAY, MCDOLE AND TRUMBULL I MOISTNESS OF SUBSOIL. 269 



whenever our data had become sufficiently numerous. This last pur- 
pose, however, was subordinated to our real object in the collection of 
samples, the determination of the relation of the extremes of moisture 
content to the hygroscopic coefficient. That our main purpose con- 
flicted more or less with the third object mentioned will be pointed 
out in a later paragraph. 

Field Notations. 

The notations as to the relative moistness of the soil, made as the 
samples were being taken, were indicated by the letters P, I, and M 
for " powder," " intermediate," and " moist," respectively. The 
sleeve was employed only where a sample could not be obtained with- 
out it ; in such cases the field notation was " P." Where the soil ad- 
hered so firmly to the bit that it could be removed from the hole by 
the ordinary auger without difficulty the condition was recorded as 
" -jyj- » Throughout the work we attempted to distinguish several in- 
termediate degrees of moistness, such as " slightly moist " and " very 
slightly moist," but as we have found no definite concordance of these 

Table i. — Relation of field notations to moisture condition in different portions 
of 12-inch sections, illustrating the difficulty of assigning a satisfactory 
notation to the foot section as a whole when an abrupt change occurs 
within it. All the borings were made in a small cornfield 
on the same day. 



Boring. 


Depth, inches. 


Field notation. 


Ratio. 


Hygroscopic co- 
efficient. 


I 


25-28 


M 


..7 






29-36 


P 


I.I 






Av. 25-36 




1-3 


a I0.0 


2 


13-14 


M 


1.8 


II.4 




15-24 


I 


i-5 


10. 




Av. 13-24 




1-5 




3 


13-21 


M 


1.8 


n-5 




22-24 


P 


1.2 


10.4 




Av. 13-24 




1.6 




4 


13-22 


M 


1.9 






23-24 


P 


1-3 






Av. 13-24 




1.8 


"11. a 


5 


13-21 


M 


i-7 






22-24 


P 


1.2 






Av. 13-24 




1.6 


a io.7 


6 


13-22 


M 


1.8 






23-24 


P 


1.2 






Av. 13-24 




1-7 


"10.9 


7 


13-18 


M 


1.8 






19.-24 


I 


i-5 






Av. 13-24 




1.6 


a 8.9 



a The two portions of the foot section had been combined before the hygro- 
scopic coefficient was determined. 



270 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



differences with the differences in the ratio of moisture content to 
hygroscopic coefficient we have, for the present purpose, included all 
of them under the designation " I." It was found that at certain 
stages of dryness the diameter, etc., of the bit employed seemed to de- 
termine whether a soil would slip off, and so be designated " P," or 
barely adhere when the auger was carefully withdrawn from the 
boring, and so be indicated as " I." The samples, except where other- 
wise indicated, were composites from 3 borings made 10 to 20 feet 
apart. 

Table 2. — Field notations on moisture conditions of samples from the three 
borings used in preparing field composites in representative fields, illustrat- 
ing the difficulty of assigning a satisfactory notation to composites 
from sections differing distinctly in moistn^ess. 



Field. 


Boring. 


Field notations 


First foot. 


Second foot. 


Third foot 


A 


I 


M 


Upper half, M ; lower half, P 




P 




2 


M 


do. 




P 




3 


M 


do. 




P 




Section 


M 


I 




P 


' B 


1 


M 


P 




P 




2 


M 


First 3 in., I; next 9 in., P 




P 




3 


M 


I 




I 




Section 


M 


I 




I 


C 


1 


M 


First 3 in., I; next 9 in., P 




P 




2 


M 


P 




P 




3 


M 


I 


First 


3 in., I; next 9 










in., 


P 




Section 


M 


I 




I 


D 


1 


M 


M 


First 


6 in., M; next 6 










in., 


I. 




2 


M 


M 


First 


6 in., I; next 6 










in., 


P. 




3 


M 


M 


First 


3 in., I, next 9 










in., 


P. 




Section 


M 


M 




I 



Where at any particular depth the soil in all three borings ap- 
peared equally moist the notation to be assigned was evident, but 
if here there were distinct differences the field description was more 
complicated. Where the upper part of a foot section was moist and 
the lower "powder," or where the reverse held true, the condition of 
!;'.'' section was recorded as "J." Where the greater part of 
thfe section wa moiti the ratio would be high, but where the opposite 
found 1 1 if ratio might prove low. This is illustrated by Table I, 
in which data arc reported from individual borings in which the foot 
■ tioi bowed a sharp break and were separated into two parts for 



ALWAY, MCDOLE AND TRUMBULL: MOISTNESS OF SUBSOIL. 2J I 

the moisture determination. The third foot of boring I showed a 
ratio of only 1.3 and the second foot of boring 4 a ratio of 1.8, al- 
though both were recorded as " I." In both cases the upper part of 
the foot was moist and the lower " powder," the ratios in the upper 
part being 1.7 and 1.9, respectively, and in the lower 1.1 and 1.3. 

The matter was still further complicated by the use of composites, 
for the reason that the depth to which the subsoil had been moistened 
or dried out varied more or less from place to place and commonly 
we did not find in any field a uniformly constant depth. The nota- 
tions on the samples taken from four fields in May, 191 2, illustrate 
this (Table 2). The samples used for the moisture determinations 
were composites of the whole foot section from 3 borings, but on the 
foot section of each boring we made separate notations. To have 
obtained all the data possible for the interpretation of these field nota- 
tions would have required separate determinations of moisture con- 
tent and hygroscopic coefficient on 10 samples from the second foot 
of Fields A, B, and C, instead of on only the three. This was a case 
where the third of the above-mentioned purposes in making field nota- 
tions had to be subordinated to the main object of the work. 

Concordance of Notations wIth Actual Moistness. 

Table 3 gives a summary of the data on over a thousand samples 
from Nebraska, 857 from the southwestern semiarid portion, and 
235 from near Lincoln in the eastern humid portion. Of the 587 
samples recorded in the field as " P " only 2 percent showed a ratio 
of 1.5 or 1.6 and none as high as 1.7, while a ratio of less than 1.4 
was shown by 95 percent of those from western and by 90 percent of 
those from eastern Nebraska. Of the samples with the notation 
" M," 87 percent of those from the western and 91 percent of those 
from the eastern part of the State showed a ratio above 1.5. Thus 
approximately 95 percent of those recorded as M had a ratio of 1.5 
or above and a similar percentage of those recorded as P had a ratio 
of less than 1.4. Thus, 1.5, 4 the computed wilting coefficient of Briggs 
and Shantz, may be regarded as the approximate dividing point, the 
soil being readily recognizable as M if it is much moister than this 
and as P when it is appreciably drier. This would lead to the con- 
clusion that when sections uniform in moistness are noted as " I " 
they should be in a moisture condition approximating the ratio 1.5. 
While all the samples from the eastern part of the State and most of 
those from the western, 709 out of 857, had a hygroscopic coefficient 

4 To be accurate, 1.47 (8, p. 65). 



272 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



of 5.0 or above, the number of the coarser-textured soils, 148, is suf- 
ficient to indicate that this property is not dependent upon a fine 
texture. 

"YYe have too few data on very fine-textured soils, those with hygro- 
scopic coefficients of 16 to 25, and on the coarser sands, to decide 
whether or not these differ from the common tillable soils represented 
in Table 3. 



Table 3. — Relation of field notations to the actual moistness, expressed as the 
ratio of moisture content to hygroscopic coefficient. 

WESTERN NEBRASKA. 





No. of 






Percentage of samples 


with ratio of 






Field notation. 


sam- 


0.9 or 
















1.7 or 




ples. 


lower. 


1.0 


1.1 


1.2 


,3 


1.4 




1.6 


higher. 


(a) Finer textured 


soils, 


hyg. coef. 5.0 


or higher. 










Powder 


472 




28 


30 


11 


5 


I 


I 


1 





Moist 


110 














3 





8 


11 


78 


Intermediate 


127 





2 


7 


23 


20 


13 


10 


8 


17 


(b) Coarser 


textured soils, hyg. 


coef. below 5. 


0. 












63 


.3 


22 


29 






8 


3 





3 


Moist 


53 













I 




8 


• 4 


76 


Intermediate 


32 








6 


i 






22 


25 


41 


(c) All soils sampled. Sum of 


a and 


b. 












Powder 


535 


22 


27 


30 


11 


5 






. 





Moist 


103 










2 


3 


8 


9 


78 


Intermediate 


159 





: 


? 


20 


17 


11 


13 


11 


20 


EASTERN NEBRASKA. 


Powder 


52 


2 


4 


28 


30 


27 


6 


2 


2 





Moist 


128 


O 











I 


1 


7 


5 


86 


Intermediate 


55 











11 


13 


22 


22 


10 


22 



W hile the decree of moistness of the subsoils corresponding to the 
field notation " P w thus appears to be similar in both the semiarid 
and humid portions of Nebraska, the frequency of occurrence of this 
dry condition shows a great difference. While in the former it ap- 

to be the prevailing condition of the subsoil, being found in 

prairies abandoned lands, alfalfa fields, grain fields after harvest, 
and at tini< even in clean cultivated orchards where the trees are still 
alive and in a fairly healthy condition, in the latter it is rarely found 
pi i!i the subsoil of well established alfalfa fields. 



ALWAY, MCDOLE AND TRUMBULL MOISTNESS OF SUBSOIL. 273 

The various levels of the subsoil have the condition M replaced by 
P only through the action of plant roots which have actually pene- 
trated into them. In the semiarid districts those portions of the sub- 
soil found to be P when a crop has died of drought, suffered serious 
injury from lack of water, or matured in a period of rather dry 
weather, but which previously, either at the time of seeding or sub- 

Table 4. — Moisture conditions in fields near Imperial, Nebr., in April, 191 1, illus- 
trating relative value of field notations compared with the actual 
determination of total moisture as an index of the available 
moisture after a prolonged drought. 



FIELD NOTATIONS. 



Depth, 
feet. 


Prairies. 


Sor- 
ghum 
stubble. 




Corn 


stubble. 




Wheat stubble. 


Winter 
wheat. 


1. 


2. 


1. 


2. 


3- 


4- 


5- 


1. 


2. 


I 


P 


P 


M 


P 


P 




I 


P 


M 


P 


P 


P 


2 


P 


P 


I 


P 


P 




I 


P 


I 


P 


P 


I 


3 


P 


P 


I 


P 


P 




I 


P 




P 


P 


I 


4 


P 


P 


I 


P 


Rock 




I 


P 


I 


P 


P 


I 


5 


P 


P 


I 


P 






I 


P 


I 


I 


P 


I 


6 


P 


P 


I 


P 






I 


P 




I 


P 


1 



TOTAL WATER. 





Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


I 


5-4 


2.3 


4.6 


6.9 


7.8 


5-2 


5-6 


8.3 


11.0 


6.8 


4.6 


2 


7-5 


2.4 


6.0 


9.4 


10.8 


9.8 


7-7 


7.0 


10.9 


8.1 


6-5 


3 


7.0 


2.2 


5-4 


10.3 


9-5 


9.1 


8.2 


5-6 


9.6 


8.0 


6.2 


4 


6.1 


2.4 


4.1 


9-7 




10. 


9.1 


10.2 


10.9 


8.9 


4-3 


5 


4.2 


2.2 


8.6 


7.6 




7.0 


4.6 


10.5 


19.8 


9.8 


3-4 


6 


4.0 


2-3 


8.6 


7-i 




5-3 


7-8 


9.9 


16.8 


7-4 


5-9 


Av. 


5-7 


2.3 


6.2 


8.5 




7-7 


7.2 


8.6 


13-2 


8.2 


5-2 










HYGROSCOPIC 


COEFFICIENTS. 










1 


6.4 


2.0 


2.9 


6.8 


7-7 


3-5 


6.0 


4-7 


13-2 


6.4 


3-2 


2 


7-9 


2.2 


4-3 


9-3 


10.2 


7-7 


7-3 


4.4 


11.5 


7-9 


4.1 


3 


8.1 


2.0 


4-3 


9.6 


8-5 


6.9 


7.6 


3-8 


9.6 


8.1 


4.1 


4 


6.8 


2.0 


3-0 


9 5 




8.0 


8.4 


6.2 


9.8 


9.2 


3-2 


5 


4.2 


2.1 


6.4 


7.2 




8-5 


4.0 


6.2 


14.9 


9.6 


2.2 


6 


3-9 


2.0 


7-1 


6.4 




3-6 


6.1 


5-3 


12.4 


7-i 


3-5 


Av. 


6.2 


2.1 


4-7 


8.1 


- 


5-8 


6.6 


5-i 


11.9 


8.1 


3-4 


RATIOS. 


I 


0.8 


I.I 


1.6 


1.0 


1.0 


i-5 


0.9 


1.8 


0.8 


1.1 


1-4 


2 


0.9 


I.I 


1.4 


1.0 


1.1 


i-3 


1.1 


1.6 


0.9 


1.0 


1.6 


3 


0.9 


I.I 


1-3 


I.I 


1.1 


1-3 


1.1 


i-5 


1.0 


1.0 


i-5 


4 


0.9 


1.2 


1.4 


1.0 




1.2 


1.1 


1.6 


1.1 


1.0 


i-3 


5 


1.0 


1.0 


1-3 


I.I 




1-3 


1.1 


1-7 


i-3 


1.0 


i-5 


6 


1.0 


I.I 


1.2 


I.I 




i-5 


1-3 


1.9 


1.4 


1.0 


1-7 


Av. 


0.9 


I.I 


1-3 


1.0 




i-3 


1.1 


i-7 


1.1 


1.0 


i-5 



2/4 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Table 5. — Moisture conditions at Imperial, Nebr., in May, 1912, illustrating 
the reliability of field notations after a wet winter following 
a very dry summer. 



FIELD NOTATIONS. 



Depth. 


Prairies. 


Corn stalks. 


feet. 




2. 


3- 


4- 


5. 


6. 


1. 


2. 


3- 


4- 


5. 


I 


M 


M 


M 


M 


M 


M 


M 


M 


M 


M 


M 


2 




M 


M 


M 


M 


M 


M 


M 




M 


M 


3 


P 


M 


M 


I 


P 


M 


P 


M 


P 


I 


M 


4 


P 


M 


M 


P 


P 


P 


P 


M 


P 


P 


I 


5 


P 


M 


M 


Rock 


Rock 


P 


P 


M 


Rock 


Rock 


I 


6 


P 


M 


M 






P 


P 


M 






I 



TOTAL WATER. 





Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


Pet. 


I 


16.0 


4.0 


7.6 


16.9 


12.2 


7-8 


17.5 


8.6 


18.3 


17-3 


14.0 


2 


16.9 


5-4 


10.3 


18.0 


14.8 


11. 1 


17.8 


8.7 


17.8 


17-5 


12.8 


3 


11. 2 


7-9 


9.4 


15-0 


9.2 


10.9 


I3-I 


8-5 


10.7 


13-6 


9-5 


4 


6-5 


6.9 


8.0 


10. 1 


6.1 


4.2 


12.0 


10. 


10.4 


10.2 


8.2 


5 


5-5 


5-9 


4-9 






3-7 


10.5 


15-6 






9.8 


6 


5-1 


6.1 


4.4 






3-6 


6.9 


21.3 


— 


— 


9.0 


Av. 


10.2 


6.0 


7-4 






6.9 


13-0 


1 2. 1 






10.5 










HYGROSCOPIC 


COEFFICIENTS. 










1 


8. 2 


1.6 


2.6 


7-1 


5-8 


3-2 


7-1 


2.4 


7.2 


7-i 


5-5 


2 


10.2 


2.6 


3-7 


7-5 


6.3 


3-2 


8-3 


2.6 


9.6 


9.2 


5-9 


3 


8.9 


1.9 


3-5 


9-7 


7-i 


5-4 


9.4 


2.6 


9-5 


8.7 


4-4 


4 


5-3 


i-5 


3-5 


9.0 


5-i 


3-7 


10.0 


3.0 


7.0 


7.0 


3-7 


5 


4-7 


i-3 


1.6 






3-4 


7.8 


5.6 






4.9 


6 


4.9 


1-3 


1-3 






3-0 


4.5 


9-3 






4.6 


Av. 


7.0 


1-7 


2.7 






3-7 


7-9 


4.2 






4.8 












RATIOS. 












1 


2.0 


2-5 


2.9 


2.4 


2.X 


2.4 


2.5 


3-6 


2-5 


2.4 


2-5 


2 


1-7 


2.1 


2.8 


2.4 


2-3 


3-5 


2.1 


3-3 


1.9 


1.9 


2.2 


3 


13 


4.2 


2-7 


i-5 


1-3 


2.0 


1.4 


3-3 


1.1 


1.6 


2.2 


4 


1.2 


4.6 


2.3 


1.1 


1.2 


1.1 


1.2 


33 


1-5 


i-5 


2.2 


5 


1.2 


4-5 


3-1 






1.1 


i-3 


2.8 






2.0 


6 


1.0 


4-7 


3-4 






1.2 


i-5 


2-3 






2.0 


Av. 


M 


3-8 


2.9 






1.9 


i-7 


3-1 






2.2 



ntly, had been found in the condition M t may be expected to 

show B ratio not far from [.I. On the other hand, if this portion of 
»il had not been in the Mate M at any time since the crop was 
planted, and 50 not in a condition to permit of root penetration, the 
Condition /' found at [tl maturity or death may correspond to a ratio 
di a- J.". 11 bcine; that induced by sonic preceding crop which 



ALWAY, MCDOLE AND TRUMBULL I MOISTNESS OF SUBSOIL. 275 

had matured under more favorable conditions of moisture supply in 
the surface layers. 



Table 6. — Moisture conditions in prairies near McCook and Wauneta, Nebr., 

in the spring of ign. 





Field notations. 


Total water. 


Depth 


McCook. 


Wauneta. 


McCook. 


Wauneta. 


















foot. 


















March 24. 


March 25. 


April 4. 


March 24. 


March 25. 


April 4. 


















Sample 1. 


Sample 2. 


Sample 1. 


Sample 2. 












Pet. 


Pet. 


Pet. 


Pet. 


I 


P 


P 


P 


P 


7.8 


7-3 


8.4 


8.8 


2 


P 


P 


P 


P 


8-3 


8.8 


7-8 


8-5 


3 


P 


P 


P 


P 


7.8 


8.7 


8.4 


9.0 


4 


P 


P 


P 


P 


7-8 


8-3 


9.8 


8.4 


5 


P 


P 


P 


P 


8.4 


8-3 


12.9 


7-5 


6 


P 


P 


I 


P 


8.9 


8-5 


10.4 


7-5 


7 


P 


P 


I 


P 


8.8 


8-5 


9-7 


7-3 


8 


P 


P 


I 


P 


9-3 


9.1 


8.7 


7-3 


9 


P 


P 




P 


9-3 


9-3 


7.6 


7-5 


10 


P 


P 


M 


P 


9-5 


9.2 


6.8 


7.6 


11 


P 


P 


M 


P 


9.6 


9.0 


6.6 


8.3 


12 


P 


P 


M 


P 


9-5 


9-3 


6.7 


8.2 


13 






M 


P 






6.8 


7-8 


14 






M 


P 






7.0 


7-3 


15 






M 


P 






8.9 


7.8 




Hygroscopic coefficient. 


Ratios. 




Pet. 


Pet. 


Pet. 


Pet. 










1 


9.6 


8.7 


8.7 


9.0 


0.8 


0.8 


1.0 


1.0 


2 


10.5 


10. 1 


9.2 


9.6 


0.8 


0.9 


0.8 


0.9 


3 


9.1 


9.6 


9.1 


10.9 


0.9 


0.9 


0.9 


0.8 


4 


8.3 


9.0 


9.2 


10. 


0.9 


0.9 


1.1 


0.8 


5 


8.1 


8.6 


9.6 


8.8 


1.0 


1.0 


1-3 


0.9 


6 


8.1 


9.0 


8-3 


7-7 


1.1 


0.9 


1.2 


1.0 


7 


8.1 


7-9 


6.6 


6.7 


1.1 


1.1 


i-5 


I.I 


8 


8-3 


8.6 


5-7 


7-3 


1.1 


1.1 


i-5 


1.0 


9 


8.1 


8.4 


4.9 


7.0 


1.1 


1.1 


i-5 


I.I 


10 


8.1 


8.4 


4.4 


7.0 


1.2 


1.1 


i-5 


I.I 


11 


8.1 


8.6 


4.0 


7-9 


1.2 


1.0 


1.6 


1.0 


12 


8.1 


8.6 


3-7 


7-4 


1.2 


1,1 


1.8 


I.I 


13 






4-3 


6.6 






1.6 


1.2 


14 






4.2 


6-5 






i-7 


I.I 


15 






5-1 


6.4 






i-7 


1.2 



Tables 4, 5, and 6 illustrate to what extent such field observation 
may afford reliable information. The notations reported in the first 
part of each table are the record entered in our field note books as 
the samples were taken, while the corresponding data on hygroscopic 
coefficient and ratio w r ere first secured some weeks, or at times many 
months, later. It will be seen that the field notes afforded immediately 
a reliable basis for further rational sampling for the purposes we had 
in view. 



2/6 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Practical Applications. 

From the preceding data it is evident that in the semiarid portions 
of Nebraska, a field examination with an auger without any collect- 
ing and drying of samples will permit a rather close estimate of the 
amount of useful moisture in the subsoil, especially where the latter 
is comparatively uniform in texture and its hygroscopicity is known. 
A still finer interpretation of the field observations is facilitated by a 
knowledge of the previous moisture condition, the character of the 
preceding weather, and the cultural history of the fields concerned. 

To illustrate the application let us assume the case of a farmer on 
loess soil near Wauneta or McCook, who, during a period such as 
that of 1909-1913, makes use of a 6-foot auger. If at planting time 
he find both soil and subsoil moist to the full depth of the auger he 
may sow his seed with the assurance that there is moisture enough, 
independent of any rainfall, to ensure the growth, or possibly even 
the maturity, of small grains. If in both soil and subsoil he find only 
" P " he will foresee that any crops then planted will be entirely de- 
pendent upon the rains to follow, while if the surface soil be M and 
the subsoil P the seed will germinate but the survival of the plants will 
be uncertain. Lastly, if the subsoil be M and only the surface layer P 
he need await only a rain sufficient to moisten this surface layer. In 
the case of certain cultivated and garden crops planted in hills he may 
add enough water to the hill to moisten a very small area from the 
surface down to the moist subsoil, so that the roots may develop down 
into the latter and draw upon it for their supply of moisture, the re- 
maining portion of the dry surface layer, that between the hills, being 
no serious handicap. 

It after harvesting the crop or at the time of plowing he find P to 
extend to a depth of 6 feet, and later, after a period of wet weather, 
find .1/ to replace it to a depth of 4 feet, he may safely assume that 
through the 4'feet the moisture content is from 2.0 to 2.4 times the 
1 y^ro-eopic coefficient. Further, as the latter value in the case of the 
loesi ubsoil averages approximately <;.o and the texture is quite uni- 
form, 1.» may even compute the weight per acre, of the maximum 
amount of available water accessible to the roots of the crops he may 
plant. 

In preparing land for an orchard or tree plantation he would prac- 
tice clean cultivation or flood it by directed storm-waters until the 
condition M is established to a depth of from 3 to 6 feet. If in a 
grove "i orchard already well established the trees ceased to thrive 
and he found / or /' a persistent condition of the subsoil he would 



ALWAY, MCDOLE AND TRUMBULL '. MOISTNESS OF SUBSOIL. 277 



prepare against the next tree-killing, dry series of years by a severe 
pruning of his trees, or still better by a thinning of the stand. 

Condition of "Powder" of Wide Occurrence in Dry-Land Regions. 
Numerous sets of samples from nonirrigable lands in New Mexico 
and Arizona indicated a similar relation of the ratio to the field nota- 
tions, P being even much more prevalent in those States than in west- 
ern Nebraska. 

In field observations unaccompanied by moisture determinations 
we have found the dry condition P to occur very widely on the drier 
lands of this country and Canada, and might specifically mention 
Akron and Parker, Colo. ; Dalhart and Amarillo, Texas ; Great Falls, 
Forsyth, and Hobson, Mont. ; Ritzville and Prosser, Wash. ; Pendle- 
ton and Echo, Oreg. ; Modesto, Clovis, and Delano, Cal. ; Indian Head 
and Moose Jaw, Sask., and Lethbridge, Strathmore, and Medicine 
Hat, Alta. 

In more humid districts, such as Minnesota, the very dry condition 
indicated by " P " appears to be even more infrequent than in eastern 
Nebraska, and hence the field observations of less importance. 

Summary. 

At the time of taking a large number of samples of soil for mois- 
ture determinations, both in semiarid southwestern Nebraska and in 
the humid eastern portion of the same State, notations were made 
as to their apparent moistness and from the correlation of these with 
data later obtained in the laboratory it has been found possible to 
give the field notations a numerical interpretation, most conveniently 
expressed as the ratio of the moisture content to the hygroscopic co- 
efficient. When the soil was too dry to be removable from the bor- 
ing by the ordinary open auger, a condition designated as " powder " 
or " P," the ratio was 1.3 or lower, whereas when it was moist 
enough to adhere well to the bit it showed a ratio of 1.5 or above. 

In the case of the semiarid soils dealt with, which had hygroscopic 
coefficients ranging from 2.0 to 14.0 and so represent the common 
tillable types, the dry condition indicated by " P " was found to be very 
common. With these the mere field examination with the ordinary 
soil auger, without any weighing or drying of samples, enables a 
quite satisfactory estimate of the moistness. Data on very fine-tex- 
tured soils and on coarse sands were too few to decide whether the 
field notations on these may be interpreted in the same manner. 

With the humid soils the dry condition represented by " P " was 



278 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

found comparatively rare, being confined chiefly to well established 
alfalfa fields, and hence in such districts the field notations have only 
a limited usefulness. 

The ordinary mineral subsoils rarely show a ratio above 2.5, roots 
appear to be unable to penetrate a soil stratum in which the ratio 
is 'below 1.5 (i. e., P), and the lower limit to which plant roots can 
reduce the subsoil moisture is approximately 1.0 or 1. 1. Therefore 
the above method of interpretation gives promise of usefulness in 
dry-land regions, not only as a convenient field aid for soil investi- 
gators and agronomists, but also as a practical method for the county 
agricultural agents and the more intelligent farmers. 

Literature Cited. 

1. Alway, F. J. Studies of soil moisture in the Great Plains region. In Jour. 

Agr. Sci., v. 4, P. 333-342. 1908. 

2. . Some soils studies in dry-land regions. U. S. Dept. Agr., Bur. 

Plant Indus. Bui. 130, p. 17-42. 1908. 

3. . Moisture studies of semiarid soils. In Rpt. 79th Meeting British 

Asso. Adv. Sci.. 1909, p. 698-699. 1910. 

4. and Clark, V. L. Use of two indirect methods for the determina- 
tion of the hygroscopic coefficients of soils. In Jour. Agr. Research, v. 
7, no. 8, p. 345-359, 1 fig. Literature cited, p. 359. 1916. 

5. , Klein, M. A., and McDole, G. R. Some notes on the direct deter- 
mination of the hygroscopic coefficient. In Jour. Agr. Research, v. 12, 
no. 4, p. 147-166. Literature cited, p. 165-166. 1917. 

6. and McDole, G. R. Relation of the water-retaining capacity of a 

soil to its hygroscopic coefficient. In Jour. Agr. Research, v. 9, no. 2, p. 
27-71, 4 figs. Literature cited, p. 70-71. 191 7. 

7. Briggs, L. J. Dry-farming investigations in the United States. In Rpt. 

84th Meeting British Asso. Adv. Sci., 191 4. p. 263-282, 7 fig., pi. 5. 191 5. 

8. and Shantz, H. L. The wilting coefficient for different plants and its 

indirect determination. U. S. Dept. Agr., Bur. Plant Indus. Bui. 230, 
83 9 fig-. 2 pi. 1912. 

9. Chilcott, E. C, and Cole, John F. Growing winter wheat on the Great 

Plains. I'. S. Dept. Agr. Farmers' Bui. 895, 12 p. 1917. 
10, TlNSLEY, J. D. ( and VERNON, J. J. Soil and soil moisture investigations in 
tli< ieason of tool. X. Mex. Agr. Expt. Sta. Bui. 38, 94 p., 1 fig., 11 pi. 
1901. 



BRYAX : HASTEXIXG GERMINATION OF BERMUDA GRASS SEED. 2JO, 



HASTENING THE GERMINATION OF BERMUDA GRASS SEED 
BY THE SULFURIC ACID TREATMENT. 1 

W. E. Bryan. 

Bermuda grass seed is one of the most difficult of all agricultural 
seeds to- germinate. The ordinary blotter method used for germi- 
nating such seeds as alfalfa, corn, beans, wheat, etc.. gives no results 
whatever in most cases. Samples sent to the Arizona station from 
time to time for germination have given rise to the necessity of ascer- 
taining a reliable method for their germination which would give 
conclusive results in a shorter period than 21 days, the time usually 
allowed for the germination of these seeds. 

The use of sulfuric acid in hastening the germination of seeds 
having hard and impervious seed coats is well known. The possi- 
bility that the slowness of the germination of Bermuda grass seed 
was likewise due to an impervious seed coat has suggested that a 
similar treatment with sulfuric acid might also hasten their germi- 
nation. To test this suggestion the following experiment was car- 
ried out. 

A sample of Bermuda grass seed from one of the local seed 
houses was obtained and 12 lots were counted out. each containing 
200 seeds. Each lot was treated with sulfuric acid for periods vary- 
ing from 5 minutes for the shortest time to 60 minutes for the 
longest time.' In treating the seeds each counted lot was placed in a 
small glass dish and enough sulfuric acid poured over to cover them. 
A glass rod was used to stir the acid so that all seeds would be quickly 
immersed. At the end of each treatment the dish containing the 
seeds and the acid was dipped into a large beaker of water, and the 
seeds washed into a cambric bag so that the acid was quickly drained 
away. The bag was then placed under a faucet and allowed to wash 
for at least 5 minutes so that all trace of the acid was removed. The 
bag was then turned wrong side out and the treated seeds were 
spread on an open blotter for germination. This is conveniently 
arranged by tying a piece of blotting paper over the top of a small 
circular glass dish about 2^ inches in diameter, the edge of the paper 
being pressed down the vertical side of the glass dish so that it reaches 
almost to the bottom and securely tied with a string. This provides 
a flat surface on top of the circular dish where the seeds are spread 

1 Contribution from the Arizona Agricultural Experiment Station. Tucson. 
Ariz. Received for publication April 8, 1918. 



280 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

out. The germinator is then placed in a larger vessel about 5 inches 
in diameter and about a half inch of water is poured into the larger 
vessel The water soon spreads to all parts of the blotter, thus keep- 
ing the seeds in contact with a free water surface without excluding 
the atmosphere from them. The large vessel is then covered and 
placed in the germinating chamber, where the temperature is kept at 
35 F. during the day and permitted to drop down to room tempera- 
ture during the night. Plate 7, figure I,, shows the blotter as used 
in these germinations. 

The number of seeds germinating in each sample was counted out 
and recorded every two days throughout the experiment. Table 1 
summarizes the results obtained and gives the total germination on 
the dates indicated at the heads of the columns. 

Table i. — Germination of Bermuda grass seed treated with sulfuric acid for 
periods of varying length. 



Total percentage germinated at each reading. 



Sample No. 


treated, 
minutes. 


Second 
day. 


Fourth 
day. 


Sixth 
day. 


Eighth 
day. 


Twentieth 
day. 


Twenty- 
second day. 


I 
2 

3 
4 
5 
6 

7 
8 

9 
10 
1 1 

12 


5 
10 

15 
20 

25 
30 

35 
40 
45 
50 
55 
60 


I 
I 

2* 
. 9 
9h 
5 

3§ 

2 










51 
68 
64 
64 
36 
27 
10 
12 
2 
I 
I 
2 


53 
71 

6S>f 

66 

46 

31* 
22! 
19 
9 

3* 

5 

5 


54 

7lJ 

71* 

69 

49 

31* 

225 

19 
14? 

6 

6* 

9t 


54 

7if 
7i* 
70 

51* 

33 

24 

19 

16^ 

6 

6* 
11 


54 
71* 
71* 
70 

S*h 

33 

24 

19 

r6| 

6| 

7 

1 1\ 



In comparison with the above, the highest germination from five 
untreated lots was 4 J / 2 percent on the tenth day, 16 percent on the 
fifteenth day. and 22 J/2 percent on the twenty-first day. 

Table 1 shows : 

1. That the Id treated for 10 minutes gives the quickest germina- 
tion in quantities sufficiently large for obtaining comparative results. 

2. That samples treated from [0 to 20 minutes give approximately 
ame germination in four days, and that these samples run fairly 

close throughout the entire germinating period. 

3. I hal 95 percent of the total germination of sample No. 2 was 
obtained at the end of the fourth flay. 

; I hat the highest percentage of germination obtained from any 
one of the untreated lots was 22]/i percent in 21 days. 



IUN01S 



Journal of the American Society of Agronomy. 



Plate 9. 




I Kcrmtula k'«» ss s, '<'I four days after setting for germination ; at 

the left, untreated, i percent germinated; al the Hulit, treated for i<> minutes. 
(# iKTccnt germinated. 



merkle: decomposition of organic matter in soils. 281 

Plate 7, figure 2, shows the results of germination after four days 
in an untreated lot and in a lot which was treated for ten minutes. 
The untreated lots germinated only 1 percent, while the lot which 
had been treated for 10 minutes had germinated 68 percent. 

It therefore seems that this method may be used to considerable 
advantage in saving time in making germination tests. As 95 per- 
cent of the seeds are germinated in four days by this treatment, and 
only 22y 2 percent were germinated in the untreated sample at the 
end of 21 days, it seems possible by this method to get a better 
estimate of the viability of the seeds in four days than with untreated 
seeds which have run throughout the entire period usually allotted by 
seed analysts to the germination of Bermuda grass seed. 



THE DECOMPOSITION OF ORGANIC MATTER IN SOILS. 1 

Fred G. Merkle. 

Introduction, 
occurrence of carbon compounds. 

Carbon compounds are universally distributed in all agricultural 
soils. They are ever being produced and consumed in the natural 
cycle of the element. The sources of gain in relation to soils are: 

1 . By bacteria ; 

2. By green plants ; 

3. By rains and snows ; 

4. Absorption of the gas ; 

5. Rise of carbon dioxide from below. 

1. Bacteria are usually regarded as liberators rather than fixers 
of the element carbon, yet species have been isolated, which perform 
the latter function. Kaserer (15) 2 demonstrated the production of 
organic matter by bacteria growing in inorganic media in an atmos- 
phere containing carbon and hydrogen. The work was confirmed by 
Nabokish and LebendefT (28), who showed the disappearance of hy- 
drogen and carbon accompanying their fixation. 

2. It is generally, not universally, assumed that green plants take 
all their carbon from the air. Thus a green crop plowed under will 
add 300 to 1,000 pounds of organic matter per acre (dry basis) or 

1 Thesis submitted for the degree of M.Sc., Massachusetts Agricultural Col- 
lege, Amherst, Mass., June, 1917. Received for publication March 3, 1918. 

2 Numbers in parentheses refer to " Literature cited," p. 300. 



282 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



approximately 0.04 percent. Green plants are, undoubtedly, the great- 
est source of gain, yet the amount is small in relation to that already 
existing in the soil. Even poor soils may contain 60,000 pounds per 

acre. 

3. Rains and snows wash C0 2 from the air, probably combined 
with NH 3 as ammonium carbonate. Schumacher (38) gives the CO, 
content of rain water as 0.3 to 1.0 volume in 1,000 of rain. Thus, a 
region having a 36-inch rainfall would annually receive from 400 to 
1.500 cubic feet or from 50 to 175 pounds of C0 2 . Such a figure 
seems very small, yet it helps to compensate the numerous losses. 

4. Soils have an absorptive power for gases, especially carbon di- 
oxide and ammonia. Ferric hydrate, alumina hydrate, humus, and 
clay appear to be the most active soil constituents as regards absorp- 
tion of C(X. Reichardt and Blumtritt (34) determined the volume 
of gas absorbed by equal volumes of various substances and the 
percentage of C0 2 contained as follows : 





Total gas absorbed 


; Percent C0 2 


Material. 


by 1 000 grams. 


volume. 




164 





Peat 




51 




14 


33 


Fe(OH) 3 


375 


70 


Fe 2 3 


39 


4 


Al(OH) 3 


69 


59 




29 


34 


Silt 


40 


32 


MgCo 3 


729 


29 


CaSO,.2H.O 


17 






The constituents found abundant in clay, viz., iron and alumina as 
hydrates, show a strong absorptive power for C0 2 . Peat is relatively 
high. Von Dobeneck (42) obtained the following results: 



Material C0 2 absorbed. 

Quartz, 100 grains 0.023 gram 

Kaolin, 100 grams 0.261 gram 

Hurtuis. 100 grams 1-773 grams 

1 • ' ' 1 » ■ grams 5.054 grams 



If we let rjuartz represent sand and kaolin clay and combine the re- 
sults of Reichardt and Blumtritt with those of von Dobeneck it is 
Nlfc to conclude thai the lOll'l absorptive capacity for CO. is largely 
due to its clay and humus content and to the state of its iron com- 
pounds. 

To show that soils actually do take on carbon by absorption the 



merkle: decomposition of organic matter in soils. 283 

results of Lemmermann (22) may be cited. He allowed a kilogram 
of soil to incubate for a period of eight weeks, determining the total 
carbon at the beginning and end of this period. An increase of 0.33 
gram was observed in one instance and 0.02 gram in another. 

5. Many carbon-containing deposits exist within the earth's crust. 
Just how much carbon may come to the surface from these deposits 
can not be determined, but it is probable that methane produced be- 
low may gradually rise to the surface and upon reaching better 
aerated conditions, be oxidized to C0 2 . The deeper soil layers con- 
tain greater quantities of C0 2 than the surface layers. Ebermayer 
(8) gives the following figures at 15 and 70 cm. respectively: 



Location. CO2 content at different depths. 

At 15 cm. At 70 cm. 

Beech woods 0.62 per cent. 1.19 per cent. 

Pine woods 1. 13 per cent. 9.39 per cent. 

Moss 1.93 per cent. 7.98 per cent. 

Sod 60 per cent. 4.13 per cent. 

Bare ground 1.19 per cent. 7.02 per cent. 



Pfeffer (29) gives the C0 2 content of the soil air at a depth of 6 
meters as 8 percent or more. 

While it is possible that the increased amount of carbon dioxide 
in the lower layers is due to the downward flow of the gas, it is more 
probable that it is diffusing up from below, in which case it would 
be an additive agent. 

SOURCES OF LOSS OF CARBON FROM SOILS. 

Soils may lose carbon (1) through leaching, (2) through evolution 
of CO.,. and (3) through possible removal by crops. 

That soils under certain conditions decrease in organic content is 
frequently observed. Walker (43) reports a decrease in humus con- 
tent on nonrotated fields as follows : 



Percent humus. 

Crop. 1895. 1905. Difference. 

Corn continuous 3-23 2.96 —0.27 

Mangels continuous 3-03 2.86 — 0.18 



Rotated fields and fields growing legumes continuously showed a 
slight gain in the ten-year period. 

Mooers, Hampton, and Hunter (27) show that only when the crop 
is removed can a decrease in humus content be expected. 

1. Loss through Leaching — Soils have a strong absorptive power 
for organic matter ; therefore, little or no carbon is lost in that form. 
The small amount of organic matter soluble in the presence of soil 



284 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



was shown by Sutton (40). He analyzed the surface water of culti- 
vated fields and found it to contain but 0.4 part of organic matter in 
100,000, a seeming insignificant amount. If organic matter were sub- 
ject to loss by leaching we would expect the subsoil of a continuously 
manured plot to contain more carbon than that of a nonmanured 
plot. Such is not the case. Dyer (7) shows that the subsoil of a plot 
manured for 50 years contains no more, even less, carbon than that 
of a plot undunged for 41 years. The difference is within the limit 
of error. 

Carbon in third 9-inch 
Condition. layer of soil. 

Dunged 9 years, undunged 41 years 0.515 per cent. 

Dunged 50 years 0.492 per cent. 

To be capable of leaching, organic matter must be soluble and when 
in solution it is easily precipitated by bases. 

Carbon as bicarbonate of lime is easily lost, as is shown by fre- 
quent analyses of drainage waters from limed fields (12). 

2. Some carbon may be lost through evolution of C0 2 , but if any x 
the amount must be slight. 

3. To say that plants may remove carbon from the soil may seem 
contrary to our teachings, yet there are numerous evidences that 
plants may derive a part, at least, of their carbon through their roots. 

It has been observed at the Rothamstead station that poor crops 
of wheat due to unfavorable climatic conditions have higher per- 
centages of ash elements than good crops. Hence minerals do not 
seem to be limiting factors. Cameron (4) uses this argument to 
prove that the use of mineral fertilizers is largely to neutralize toxic 
substances, but it could be used equally well to show that the syn- 
thesis of organic matter as well as the assimilation of minerals is an 
important factor in plant growth. 

To -how tin- value of organic matter in aqueous extracts of poor 
soils the Bureau of Soils, according to Cameron (5), used a manure 
extract a> follows: One portion of the extract was evaporated 
and ignited to destroy the organic matter. The other part was used 
without ignition. The solution to which the Uliignited manure ex- 
tract was added gave a far superior growth. Cameron attributes the 
value of the organic matter in the extract to its probable absorbent 
action on toxic substances, but it is also probable that the plants ab- 
sorbed certain organic nutrients from it. 

Gardnci ( 11 ) determined the effect of many substances, mineral 
and organic, on transpirat ion and upon the amount of green matter 



merkle: decomposition of organic matter in soils. 285 

produced per unit of water transpired. The following figures give 
the summarized results of many trials: 



Growth due to Growth per unit of Transpiration per 

Material added. fertilizer. water transpired. unit of growth. 

Nothing, check 100 100 100 

P 104 103 97.0 

K 113 107 93.6 

KP 118 108 92.6 

Lime 127 103 97.0 

N 145 116 86.2 

NP 144 119 84.0 

NPK 152 123 81.3 

NK 154 125 80.0 

NPK lime 173 129 77.5 

Manure * 193 135 74.0 

Clover and lime 197 143 69.9 



It will be noted that the last two treatments, which are organic, 
not only gave the greatest growth, but gave the greatest growth per 
unit of water transpired. This work was done with soil solutions so 
the effects of the organic matter can not be due to its action on the 
physical condition of the soil, nor to its solvent action upon minerals. 
It is fair to conclude that the presence of carbon in the soil solution 
decreases the transpiration necessary to produce a unit of dry matter, 
a strong indication that plants may assimilate carbon through their 
roots. 

Quarrie (32) reports large increases in garden crops through the 
application of carbon dioxide to the soil through pipes. Bornemann 
(2) reports like results with spinach. Mitscherlich (24), on the other 
hand, obtained no increase from the application of water saturated 
with C0 2 . The possibility of adding an excess of water or of gas 
renders the results inconclusive. We know that in ordinary practice 
C0 2 producing materials are seldom injurious. 

De Saussure (6) compared the growth of plants in pure water 
with water containing one-fourth its volume of carbon dioxide and 
found that the carbonated water was injurious to growth in the early 
stages, but not so later in the life of the plant. At the conclusion of 
the experiment the plants grown in the carbonated water weighed 
46.4 grams, while those growing in pure water weighed 45.5 grams. 

Hellreigel and Wilfarth (13), Franke (9), Berthelot (1), and 
Schlossing and Laurent (37) all report the utilization of organic ni- 
trogen by green plants. Schreiner and his associates (39) have iso- 
lated creatinine, an organic nitrogen compound, from soils and proved 
its beneficial action upon plant growth. Lefevre (18) grew plants in 
an artificial soil made from sand and moss, supplied with amids and 
sterilized so that further oxidation of these compounds would be 



2S6 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

avoided. The entire plant was enclosed in an atmosphere freed from 
carbon dioxide. Under such conditions it is evident that any growth 
must result from the assimilation of the amids, Lefevre obtained 
normal growth and concludes that: I. In a soil supplied with amids 
one may develop green plants without carbon dioxide. 2. The 
growth thus produced is a real synthesis, not a pousee aqueuse (19). 
3. Without light, synthesis from amids is impossible (20). 

So much for nitrogenous organic substances. Molliard (25), 
using glucose, and Laurent (17) and Knudson (16), using other car- 
bohydrates, have shown that plants assimilate sugars and that these 
sugars are used to synthesize dry matter. 

Ravin (33) compared the effects of organic acids with their acid 
and neutral salts and concluded that such organic acids as malic, tar- 
taric, citric, succinic, and oxalic may be assimilated by plants and 
further that these organic acids are more nutritive than their corre- 
sponding neutral salts or acid salts. 

So far we have considered the assimilation of carbon from ma- 
terials of known composition, namely, C0 2 , amids, carbohydrates, 
and organic acids. Molliard (26), to put the matter on a more prac- 
tical basis, experimented with humus extracted from soil. The work 
was carried on under sterile conditions, but it was impossible to pre- 
vent entirely the evolution of CCX ; therefore, definite conclusions 
can not be drawn. 

The most conclusive proof that green plants can take up carbon 
compounds through their roots is their growth with the foliage en- 
closed in an atmosphere entirely devoid of carbon dioxide. Pollacii 
(30) grew plants in a culture bottle within a large receptacle, each 
being provided with tubes so that the water or air in each may be re- 
newed and controlled independently of the other. The plants were 
^aled into the stopper with wax. By adding C0 2 to the nutrient 
solution and excluding it from the aerial portions of the plant he has 
Successfully grown pknta and even revived the chlorophyl in etiolated 
leaves. 

\ ■<•-: evidence in the foregoing pages it may be concluded 

that k'reen pkmts can, and probably do, take carbon through their 
roots. Just what form or wlia! proportion of the total carbon in the 
plant lln- may be can not be stated, but the fact itself is enough to 
make u« turn our attention to the soil organic matter. 



MERKLF I DECOMPOSITION OF ORGANIC MATTER IN SOILS. 287 
DECOMPOSITION OF ORGANIC MATTER. 

Hopkins (14) states that "It is the decay of organic matter and 
not the mere presence of it that gives life to the soil. Partially de- 
cayed peat produces no such effect upon the productive power of the 
soil as follows the use of farm manures or clover residues." Lohnis 
(23) declares that the organic matter is the life of the soil and upon 
its decay depends the fertility of the soil. 

Realizing the importance of organic matter and its decomposition 
with reference to soil fertility, many investigations have been made 
to demonstrate the rate of decay and the factors influencing it. 

Van Suchtelen (41) has used the rate of decay, measured by car- 
bon-dioxide production, as a measure of bacterial activity. This 
method recognizes C0 2 as the ultimate and most representative end 
product of decay. He showed the influence of moisture and of frost, 
the effect of soluble sugars and of salts on bacterial activity. His re- 
sults showing the action of fertilizers on the rate of decay are closely 
related to our subject and will be reported. He mixed the materials 
in 6 kg. of soil and determined the amount of carbon dioxide pro- 
duced in 12 hours. His results were as follows: 



Materials used CO2 produced. 

6 kg. soil, no addition 145 mg. 

6 kg. soil + 90 gr. MgSOJLO 408 mg. 

6 kg. soil -f- 6 gr. CaO 62 mg. 

6 kg. soil + 30 gr. (NH 4 ) 2 S0 4 864 mg. 

6 kg. soil-f- 6 gr. superphosphate 306 mg. 



The increases from applied materials are quite large with the ex- 
ception of lime, which has evidently absorbed the gas produced. One 
function of fertilizers may be to hasten the decay of organic matter. 

Lemmermann (21) and associates worked with the influence of 
lime compounds on decay. They compared the oxide and carbonate. 
They found that C0 2 production could not be taken as a measure of 
bacterial action with lime, because the oxide absorbed and the car- 
bonate gave up C0 2 . To offset the difficulty they carried on balance 
experiments in which the total carbon was determined before and 
after the incubation period, which lasted eight weeks. Their exper- 
iments show that (a) lime hastens decay, (b) kainit and a mixture 
of kainit and superphosphate do not increase decay, and (c) dry 
organic matter decays as rapidly as the same material fresh. 

Potter and Snyder (31) report some work along this line. In 
their experiments the soil was placed in pots under bell jars and the 
C0 2 evolved was measured by drawing air over, not through, the 
soil. Their observations will be mentioned later. 



288 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Fred and Hart (10) showed that sulfate of ammonia, sulfate of 
potash, and phosphates increased the carbon-dioxide production, the 
first named to a marked degree. 

Russell (36) measures oxidation by determining the oxygen ab- 
sorbed rather than the C0 2 produced. Either method should give 
about the same results, for many analyses show that a high oxygen 
content of soil air is accompanied by a low C0 2 content and vice 
versa. In other words, the sum of the oxygen and carbon dioxide is 
nearly constant. Russell's method is to place the soil in a flask, con- 
nected on one side to a KOH flask and on the other side to a mercury 
tube. The KOH absorbs any C0 2 evolved and the rise of mercury 
in the other arm indicates the oxygen absorbed. He determined the 
oxidation of many soils by this method and concluded that in different 
soils of the same type the rate of oxidation varies in the same way 
as the fertility and may be used as a measure of it. This, if true, is 
important, for we have no other laboratory method of determining 
the relative fertility of soils. 

Experimental. 

The work of previous investigators indicates oxidation to be a 
measure of fertility in soils ; hence, the rate of oxidation of organic 
additions should be a measure of their effectiveness. For the pur- 
pose of comparing organic materials ordinarily added to the soil the 
following series of experiments were planned. 

For determining the rate of oxidation quart milk bottles were used. 
They were fitted with 2-holed rubber stoppers, one hole carrying a 
short glass tube while the other carried a tube reaching to the bottom 
of the bottle. Both tubes were fitted with short rubber connections 
Stopped with glass plugs. Two hundred grams of washed gravel 
were placed in the bottom of the bottle to facilitate aeration and 
afford a space for the excess CO.. The organic substance used in 
•1 ■■ '< ~\ w:\-< thoroughly mixed with 300 grams of moist soil (25 per- 
cent rater) and placed Oil top of the gravel. The soil was mod- 
erately compacted by tamping. 

The joil used irai a fine, sandy loam of alluvial formation which 

had been under cultivation for many years. It was stored in covered 
ash bar:' 1 and not allowed to dry out, so the original bacterial flora 
uflinenl for the work. To make sure of this one bottle was 
Inoculated with TO c.c. of a manure suspension. This bottle gave the 
same amount of CO, as the nil inoculated one after the first week of 
incubation, Oiowing that there was no deficiency of organisms. 



merkle: decomposition of organic matter in soils. 289 




Fig. 41. Apparatus used in the ex- 
periments. See description in text. 



The rate of oxidation was determined by measuring the amount of 
C0 2 produced each week, as follows. The rubber connections were 
closed with pinch cocks, the glass plugs removed and the bottles con- 
nected with the absorption bottles as shown in figure 41. First is an 
absorption bottle (A) containing NaOH to free the incoming air of 
C0 2 . Next is the incubation bottle 
(B) with its outlet tube reaching 
the bottom to make sure of com- 
plete removal of the C0 2 produced. 
The absorption apparatus (C) was 
devised to take the place of a Reis- 
set (35) absorption tower. The 
tower (D) is an ordinary 100 c.c. 
pipette filled with broken glass or 
beads to increase the absorption 
surface. The pipette is connected 
with a Chapman filter pump. It 
was found that a rapid stream of 
air could be drawn through this 
tower without danger of incom- 
plete absorption, and also that four minutes of strong aspiration was 
sufficient to remove all C0 2 from the generating flask. 

Each bottle was aspirated once a week, using 50 c.c. of N/2NaOH 
as the absorbent. The C0 2 was determined by the double titra- 
tion method (3). A 10 c.c. aliquot of the carbonated soda is titrated 
with phenolphthalein against HQ, first using normal acid until near 
the neutral point. Neutralization is completed with N/10 acid. This 
marks the conversion of carbonate to bicarbonate, neutral to phe- 
nolphthalein. 

Na 2 C0 3 + HC1 + phenolphthalein NaHC0 3 + NaCl. 

The amount of acid needed to make this change need not be known, 
nor is it necessary to know the normality of the alkali used. 

Methyl orange is now added and N/10HCI run in drop by drop 
till the neutral point is reached. The exact amount is recorded and 
is equivalent to the C0 2 contained. 

NaHC0 3 + HC1 + methyl orange -> NaCl + H 2 + C0 2 . 

One cubic centimeter of N/10HCI equals 4.4 milligrams of C0 2 . 
Cochineal gives about the same results as methyl orange, but the 
latter was used throughout this work. 



29O JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

EXPERIMENT I, LEGUME FODDERS. 

Soybeans, alfalfa, and red clover were used in this experiment. 
The plants were cut off at the surface of the ground when in full 
bloom or as near that stage as possible. They were dried, slowly at 
first and later in the oven. When dry they were ground and re- 
ground until all the material would pass through a 2-mm. sieve. Fif- 
teen grams were mixed with 300 grams of moist loam, placed in the 
incubation bottles on top of a layer of gravel and slightly compacted. 
The bottles were stoppered and the outlet tubes closed with glass 
plugs. They were allowed to incubate in the dark at room tem- 
perature, the C0 2 produced being measured weekly (usually) in the 
manner just described. An untreated soil served as a check for all 
the following experiments. The results are shown in Table 1. 



Table i. — Milligrams of CO2 given off from untreated soil and from soil to 
which various legume fodders were added. 



Date. 


Loam 300 gr. 
untreated. 


Loam 300 gr. -f soy- 
bean fodder 15 gr. 


Loam 300 gr. + al- 
falfa fodder 15 gr. 


Loam 300 gr. 4-red 
clover fodder 15 gr. 


Nov. 15 


35-2 


475-2 


444-4 


426.8 




33-0 


385-0 


336.6 


325-6 


Nov. 29 


50.0 


211. 


242.O 


154-4 


Dec. 6 


52.0 


213.4 


281.6 


195-8 


Dec. 13 


37-4 


167.4 


200.2 


182.6 


Dec. 21 


48.4 


206.8 


228.8 


206.8 


Dec. 28 


50.6 


237.6 


193-6 


162.8 


Jan. 10 


41.8 


191. 4 


189.2 


132.0 


Jan. 17 


37-4 


195-8 


l67.2 


143-0 


Jan. 28 


28.6 


167.2 


158.4 


162.0 


Feb. 6 


35-2 


160.6 


165.O 


169.4 




449-6 


261 1.4 


2607.0 


2261.2 



The results shown in Table I are plotted in figure 42. They show 
thai a rapid production of C0 2 takes place the first two weeks after 
a legume fodder starts to decay, and that after the second week they 
Settle Mown to a steady rate of decomposition. Apparently red clover 
decays a little slower than the other fodders, but there is no great 
difference between them. 

'I here are po Slbilitiea of errors in the aspiration of the gas, but the 
llaritiefl In the curves arc noi due to these. Temperature changes 
affect all alike, hence the general tendency is for all to rise and fall 
at the same period, though not always in the same degree. The uni- 
formity of the check indicates the accuracy of the method. Dupli- 
cates were run in the early pari of the experiment but the close 
men! 1 < merl to justify dropping them to save work. 



merkle: decomposition of organic matter in soils. 291 



Humus Production. — Equally important as the rate of oxidation is 
the humus produced. A substance may oxidize very rapidly, as, for 
example, sugars, and still not increase the humus content noticeably. 
Such substances would be of questionable value as regards the phys- 
ical improvement of the soil. Unpublished work by the writer shows 
that sugars break up very rapidly in the soil and are nearly com- 

























































































































































































































































































































































































































































































































































































































































































































































































i 






ft 


















































































— — f 






























ii 




\- 
























T" 
































































\ 














Jt 




d 


a 
































































































































































































































































































c 


m 




i 







GOO 

soa 
+00 

300 
ZOO 

/oo 



!? 8 9 ^ 5 «5 S «5 3 ^ 



Fig. 42. Graph showing C0 2 given off during various periods from untreated 
soil and from soil to which various legumes were added. (Data from Table 1.) 



pletely oxidized within a week or two. Lactose, maltose, saccharose, 
dextrose, and fructose run about the same. Sugar beets (figure 43) 
in the early stages of decay show the effect of their sugar, but later 
gave about the same results as rape and swedes. 

The materials used in the C0 2 production experiments, having 
been allowed to incubate from November 8 to February 19, were 
removed, dried, and their humus content determined by the official 
method. The results are recorded below, together with the total 
C0 2 production for comparison. 



292 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Treatment. Humus, percent. Total CO2, eg. 

Soil, no treatment 2.96 44 

Soil + alfalfa 3.43 260 

Soil -f red clover 3.29 220 

Soil -f- soybeans 3.28 261 

The figures indicate that there is little choice between the legumes 
in decay and humification. 



EXPERIMENT 2, ROOT CROPS AND RAPE. 

Root crops and rape were used to compare readily decomposable 
carbohydrates, as found in plants, with more inert materials. For 
this purpose sugar-beet roots, swede or rutabaga roots, and rape tops 
were used. All of these contain some form of stored food, sugar or 
starch. The plants were taken from the field, air dried, then oven 
dried, and ground fine enough to pass a 2-mm. sieve. Fifteen grams 
of each were mixed with 200 grams of moist soil and placed in incu- 
bation bottles as previously described. Determinations of C0 2 pro- 
duced were made weekly. The results are shown in Table 2 and are 
also shown graphically in figure 43. 



Table 2. — Milligrams of C0 2 given off from untreated soil and from soil to 
which sugar beets, rutabagas, and rape were added. 



Date. 


Loam 300 gr. 
untreated. 


Loam 300 gr. + 
sugar beets 15 gr. 


Loam 300 gr. + 
rutabagas 15 gr. 


Loam 300 gr. -(- 
rape 15 gr. 


Nov. 15 


35-2 


550.0 


464.2 


4OO.4 




33-0 


708.4 


484.O 


396.0 


Nov. 29 


50.0 


213.4 


261.8 


231.0 


Dec. 6 


52.0 


235-8 


226.6 


244.2 


Dec. 13 


37-4 


171.6 


162.8 


165.O 


Dec. 21 


48.4 


132.0 


189.2 


206.8 


Dec. 28 


50.6 


160.0 


165.O 


182.6 


Jan. 10 


41.8 


125.4 


147.4 


158.4 


Jan. 17 


37.4 


103.4 


132.0 


1 1 0.0 


Jan. 28 


28.6 


zz8.8 


II4.4 


149.6 


Feb. 6 


352 


106.8 


I40.8 


150.2 




a 19.6 


2625.6 


2488.2 


240O.2 



Sugar beets, aa might be expected, show rapid decay at the start 
but the BUgar is all oxidized in two weeks, after which time the or- 
ganic matter in them is no more decomposable than that of other ma- 
terials. Rutabaga contain but little sugar and decay no faster than 
legume fodders. Rape LI BlowesI at first but as time goes on it ex- 
ceeds the others. 

( omparing the legumes with roots wc find that the former are 



merkle: decomposition of organic matter in soils. 293 



more readily oxidized as time goes on, that is, after the sugar in the 
roots is broken down. 































































































































































































































































































































































































































































































































































































































































































































































/ 




t 




















































/ 






















































































































1 
























































































— 














































—\ 


>M 


re 


d 


Si 




























































J 


?£ 








































S 


















M 
































































































































— 
















































C 


hi 


ft 









7O0 
600 

500 
400 

300 
2.00 

too 





v> 5 
^ 



\ 00 q 
<M <\ 

a Or ^ 



^ ?? ^> 



Fig. 43. Graph showing CO2 given off during various periods from untreated 
soil and from soil to which sugar beets, rutabagas and rape were added. (Data 
from Table 2.) 



The results of the humus determinations are as follows : 

Treatment. Humus, percent. Total CO2, eg. 

Soil, no treatment 2.96 44 

Soil -j- swedes 3.56 248 

Soil + sugar beets 3.28 262 

Soil + rape 3.24 240 

The difference as shown by the humus figures seems the more rep- 
resentative, since the higher C0 2 production for sugar beets is due 
to the sugar. Rape falls in third place in both instances. 



294 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



EXPERIMENT 3, LITTERS. 

The materials listed below find their way into the soil through 
natural agencies or as litters and were selected with the expectation 
of obtaining large differences. It was thought that pine needles might 
even lower the bacterial activity, at least for a time. 

Pine needles, oak leaves, and maple leaves were picked while still 



































































































































































































































































































































































































































































r 






































































































































































































































































































































































































































5 


hi 


























































VI 




















































H 








r 4 


a. 














































C 




Y' 


















































7 


V? 




ft 
i 


& 




IP 


r e 




















































Zc 


Yi 


'S 


















































































































C 


he 


cA 









I 



500 



zoo 

ZOO 

/oc 



; Graph ihowiiiK ( ( > kivcii off during weekly periods by untreated soil 
fl I to ,vliicli varion liltn were added. (Data from Table ) 



00 irx 



n, air dried and later nven dried. White pine shavings, as used 
for litter, were oven dried. Kach substance was ground and sieved. 
Fifteen grams were used in eaeli case. The CC) 2 determinations are 
diown in I aide 3 and graphically in figure 44. 

White pin'- shavings stand out as a striking example of an inert 
ttb t&flCe, being lowest and slow ! in CO., production. Maple leaves 
a more uniform decline than anything else. 



MERKLE I DECOMPOSITION OF ORGANIC MATTER IN SOILS. 



295 



The litters in general, as might be expected, are not as rapidly de- 
composed as either legumes or root crops and suggest the importance 
of nitrogen as an aid to oxidation, as those materials which are low 
in nitrogen are slow to oxidize. This latter statement applies to the 
later stages of decomposition. 



Table 3. — Milligrams of C0 2 given off from untreated soil and from soil to 
which various litters were added. 



Date. 


Loam 300 gr. 
untreated. 


Loam 300 gr. -f- 
shavings 15 gr. 


Loam 300 gr. + 
maple leaves 
15 gr. 


Loam 300 gr. + 
oak leaves 15 gr. 


Loam 300 gr. -f- 
pine needles 15 gr. 




35.2 


257.4 


275.6 


338.8 


343-2 




33-0 


257.4 


250.8 


303.6 


259.6 


Nov. 29 ...... 


50.0 


118. 8 


224.4 


182.6 


224.4 


Dec. 6 


52.0 


156.2 


211. 2 


178.2 


167.2 


Dec. 13 


37-4 


103.4 


206.8 


187.O 


I93-I 


Dec. 21 


48.4 


88.0 


184.8 


180.4 


184.8 


Dec. 28 


50.4 


105.6 


l67.2 


162.8 


178.2 


Jan. 10 


41.8 


101.2 


I7I.6 


147-4 


151-8 


Jan. 17 


37-4 


77.0 


149-6 


158.4 


136.4 


Jan. 28 


28.6 


132.0 


143.0 


149.6 


134.2 


Feb. 6 


35-2 


74-8 


160.6 


116.6 


110.0 


Totals .... 


449-6 


1471.8 


S 2145.0 


2105.4 


2083.4 



The rate of oxidation, as measured by humus production and C0 2 
production, follow the same order, namely, (i) maple leaves, (2) 
oak leaves, (3) pine needles, and (4) pine shavings. It should be 
noted that the shavings after having been in the soil for three or four 
months did not increase the percentage of humus ; in fact, they low- 
ered it slightly. The data are shown below. 

Material used. Humus, percent. Total CO2, eg. 

Soil alone 2.96 44 

Soil -f- maple leaves 3.34 2I 4 

Soil + oak leaves 3-!8 210 

Soil + pine needles 3.07 208 

Soil + shavings '. 2.91 147 

EXPERIMENT 4, CEREALS AND BUCKWHEAT. 

Barley, oats, and buckwheat were used in this experiment because 
good samples of them were available. Barley and buckwheat are 
quite frequently plowed under as green-manure crops, which is not 
true of oats. Plants that were half matured were dried, ground, and 
mixed with the moist loam. The rate of oxidation is shown in Table 
4. The data are also shown graphically in figure 45. Little or no 
consistent variation occurs. Buckwheat appears to be the most inert. 



296 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Table 4. — Milligrams of COz given off from untreated soil and from soil to 
which oats, barley, and buckwheat were added. 



Date. 


Loam 300 gr.. 


Loam 


Loam 3 g**. 


Loam 300 gr. 


untreated. 


oats 15 gr. 


+ buckwheat 15 gr. 


+ barley 15 gr. 


Nov. 15 


35-2 


349-8 


442.2 


428.8 




33-0 


338.8 


283.8 


380.6 




50.0 


242.0 


176.0 


341-0 




.52.0 


239.8 


T 80 A 

Io2.0 


2o 1 .0 


Dec. 13 


37-4 


176.0 


I84.8 


I9I.4 




48.4 


253.0 


138.6 


195-8 


Dec. 28 


50.6 


184.8 


125-4 


176.0 




41.8 


189.2 


162.8 


138.6 


Jan. 17 


37-4 


176.0 


160.6 


134.2 


Jan. 28 


28.6 


184.8 


158.4 


II8.8 


Feb. 6 


35-2 


147-4 


136.4 


123.0 


Totals 


449.6 


2481.6 


2151.6 


25J9-2 































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































-• 






























o\ 




s 






















































r 


















































I 




it 


A 


14 


ft 


e< 


it 




























s 






/ 




s 




A 


b 


& 


M 
















































































































































































c 






H 







l| II 1 4 \ 1 4 \ I 

I r Graph showing COi given off during weekly periods from untreated 
•Oil tad from soil to wln< h Oftl . ImiIc.v, and btlCkwheal were added. (Data 
from Tabic 4.) 



MERKLE : DECOMPOSITION OF ORGANIC MATTER IN SOILS. 297 



The data on humus production of oats, barley, and buckwheat are 
as follows : 

Material. Humus, percent. Total C0 2 , eg. 

Soil alone 2.96 44 

Soil 4- oat fodder 3.18 248 

Soil -f- barley fodder 3.10 252 

Soil + buckwheat fodder 2.99 208 

The humus production of oats and barley is very nearly the same, 
the variation being within the limits of error. Buckwheat seems to 
be a very inert substance, increasing the percentage of humus almost 
nil, while the total C0 2 given off in three months is considerably 
lower than than the other materials. 

GENERAL OBSERVATIONS. 

Before the experiment was started it was expected that a wide 
variation in the rate of decomposition would be shown. Wollney 
(44) states that " Legume straws containing a high nitrogen content 



Swecfes i 
MapleLeaves^L 



3-43 



3.28 



3.1 85 



Be c?C lover I 3 . 29 

3.28 5 

SucjarJZeets z 
Rape c 
Oats c 

Ba.r!ey t 
PrneWeecJles c 
Buckwheat c 
Shav/nys c 



2.99 



CfiecK i 2.97 i 

Fig. 46. Graph showing humus production from all materials used in 
the experiments. 

are easily decomposed, grain straws are more resistant, while leaves 
and needles are still more so." The results show this to be true, but 
the difference is not as marked as might be expected. That white 
pine shavings should increase the C0 2 production as much as they did 



2 98 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

is peculiar so it seems that the increased aeration afforded by the 
loose material has had some effect in causing a greater recovery of 

CO,. 

It should be remembered that all substances were dried before 
using, which may account for the uniformity of the results, although 
Lemmermann (21) found no difference between green and dry 
lucern. It would be nearly impossible to obtain, at the same time, all 
of the materials at the proper stage of growth and normal moisture 
content. To place everything on the same basis it seemed advisable 
to dry each in the same degree. 

For the sake of comparison the humus production of all the ma- 
terials is given in figure 46. It is believed that these results fairly 
represent the availability of the substances used. 

Action of Fertilizers on Decomposition. 
A second series of experiments was run along the same period as 
those just cited in an effort to determine whether or not fertilizer 
materials increased the rate of decomposition. The same form of 
apparatus was used. Fifteen grams of soybean fodder and one gram 
of the fertilizer to be tried out were added to each flask. The results 
are shown in Table 5. 



Table 5. — Milligrams of CO- given off by soybean fodder to which various 
fertilizing 7naterials had been added. 



Date. 


Soybeans 
alone. 


Soybeans + 
sulfate of 
ammonia. 


Soybeans + 
nitrate of 
soda. 


Soybeans -f 
ammonium 
phosphate. 


Soybeans + 

calcium 
cyanamid. 


Soybeans -f- 

acid 
phosphate. 


Nov. 15 


475-2 


437-8 


431.2 


289.4 


244.2 


411. 2 




3850 


365.2 


545-6 


374-0 


330.0 


374-0 


Nov. 29 


21 1.2 


220.0 


279-4 


136.4 


195-8 


195.8 


Dec. 6 


213-4 


228.8 


297.0 


198.0 


206.8 


209.0 


Dec. 13 


169.4 


158.4 


211. 2 


180.4 


32I.O 


180.4 


Dec. 28 


237.6 


187.O 


235-4 


224.4 


182.6 


195-8 


Jan. 17 


195-8 


138.6 


158.4 


173-8 


283.8 


1 7 1. 6 


Jan. 28 


1 17.2 


145-2 


143.0 


154.0 


162.8 


132.0 


Totals 


2054.8 


1881.0 


2301.2 


1830.4 


1927.O 


1870.0 



l>»te. 


Soybeans + 
raw bone; 


Soybeans -f 
basic slaR. 


Soybeans + sul- 
fate of potash 


Soybeans -f 
kainit . 


Soybeans + mu- 
riate of potash. 




296.O 


4U.6 


446.6 


352.0 


407.0 




341.0 


389.4 


409.2 


255-2 


191.4 




228.8 


193-6 


209.2 


224.4 


204.6 


Dec. 6 


.."><,. 


283.8 


184.O 


202.4 


204.6 


Dec. 13 


[6p«4 


2 59.0 


191.4 


158.4 


136.4 


Dec. 38 


220.0 


242.I 


217.8 


228.8 


I7I.6 


Jan. 17 


167.2 


187.O 


147.4 


167.2 


134-2 


Jan. 38 


156.2 


173-8 


I54.0 


112.2 


1 12.2 


Total* 


1887.4 


3I4I.4 


1952.8 


[700*6 


1 563.0 



MERKLE : DECOMPOSITION OF ORGANIC MATTER IN SOILS. 299 

The results show that but two of the fertilizer materials tried out 
increase the rate of decay; these are nitrate of soda and basic slag. 
The others show but little effect with the exception of kainit and 
muriate of potash, which decrease the rate quite markedly. The re- 
sults with kainit agree with the carbon balance experiments of Lem- 
mermann, previously mentioned. 

Calcium cyanamid contains carbon, so it is not fair to draw any 
conclusions regarding its effect on organic decay as measured by C0 2 
production. However, it appears to be toxic to soil bacteria as is 
shown by the markedly lowered production during the first two 
weeks. This toxic action seems to last but one week, agreeing with 
the recommendations of Brooks, Schneidewand, and others that the 
material be applied a week or two before planting time. 

More experimental work of this kind has been done with sulfate 
of ammonia than any other fertilizer and contradictory results have 
been obtained. Van Suchtelen, using a light application of sulfate 
of ammonia and measuring the C0 2 for a very short period (12 
hours), obtained much more gas from the treated soil. Fred and 
Hart (10) made determinations at 2-day periods and, while an in- 
crease over the check is shown, it is not nearly as great as the above. 
Potter and Snyder (31) found a slight decrease in C0 2 production 
from the use of sulfate of ammonia, as did the writer. The results 
of the last two experiments are not entirely contradictory to the 
former, for the time factor enters. It seems that the immediate 
effect of the salt is to increase or stimulate bacterial action, but it is 
not lasting. The results obtained here, as well as those of Potter 
and Snyder, represent a length of time equivalent to a growing sea- 
son and for that reason should be of more practical value. 

The residues from the oxidation experiments were dried and their 
humus content determined. The results were as follows : 



Treatment. Percentage of humus. 

Soybeans 15 gr. alone 3-285 

Soybeans 15 gr. -f- kainit, 1 gr 3-225 

Soybeans 15 gr. -f- raw ground bone, 1 gr 3- IQ o 

Soybeans 15 gr. -}- muriate of potash, 1 gr 3-i8o 

Soybeans 15 gr. -f- sulfate of ammonia, 1 gr 3-175 

Soybeans 15 gr. -f acid phosphate, 1 gr 3-155 

Soybeans 15 gr. -f- calcium cyanamid, 1 gr 3-130 

Soybeans 15 gr. -f sulfate of potash, 1 gr 3.035 

Soybeans 15- gr. + ammonium phosphate, 1 gr 3.000 

Soybeans 15 gr. -f rock phosphate, 1 gr 2.990 

Soybeans 15 gr. -f- basic slag, 1 gr 2.970 

Soybeans 15 gr. -f nitrate of soda, 1 gr 2.865 



300 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Those materials which markedly depressed the production of C0 2 , 
viz., kainit and muriate of potash, caused the least loss in humus. 
This is shown by the relatively high humus content in the jars treated 
with those substances. On the other hand the materials which in- 
creased the production of C0 2 , viz., slag and nitrate of soda, have 
markedly lowered the humus content. Considering this one may infer 
that fertilizers act upon the soil humus and not upon the crude or- 
ganic matter. One would expect the continued use of materials like 
nitrate of soda to cause a rapid depletion of the soil's humus content. 

Summary and Conclusions. 

1. The legumes which are high in nitrogen show a more rapid rate 
of decay than straws and litters which are low in nitrogen. Nitrogen, 
then, seems to influence decomposition. 

2. On farms where animal manures are not available the choice of 
green manures and cover crops is important. The results indicate 
that legumes would be most desirable on such farms. 

3. Cyanamid appears to be toxic to soil bacteria, or at least arrests 
the decay of organic matter for two weeks after application. 

4. Commercial fertilizers apparently act upon soil humus, decom- 
posing it quite rapidly. They apparently do not act upon crude or- 
ganic matter in the same way. 

Literature Cited. 

Editor's Note. — The citations of literature as originally made by the author 
in the preparation of his thesis were not as complete as those usually printed in 
The Journal of the American Society of Agronomy, and unfortunately he 
is not now able to complete some of them. It is hoped that they will be suffi- 
ciently complete to be useful to the reader. 

1. BotTHKLOT, . In Ann. Chem. Phys., 13: 5. 

2. Borneman, . In Dcutsch. Landw. Gesell., v. 28, no. 31, p. 443. 1913. 

3. BlOWll EL, and Escomb, . Researches on some of the physiological 

processes of green leaves, with special reference to the interchange of 
between the leaf and its surroundings. In Proc. Roy. Soc, 76: 

29. 1905. 

4. Ca MHO*, P. K. The soil solution, p. 14. Chemical Pub. Co., Easton, Pa. 

19". 

5. . Op. cit., p. 85. 

f > I" Sai 1 hi:, Theoi). Recherchcs chemiques snr la vegetation, p. 27, 28. 
Paris, 1804. 

7 !>••'»' Mi mm.. Results of investigations on the k'othamstcd soils. U. S. 

hepi v. ■! ( Mi- 1 Expt Sta. BuL 106, p. 30. 
8. Kbkjim ayp.r, K. Untcrsuchungen ubcr die bedeutung des Humus als Boden- 



merkle: decomposition of organic matter in soils. 301 

bestantheil and iiber den Einfluss des Waldes, verscheidener Bodenarten 
und Bodendeken auf die Zusammenzetzung der Bodenluft. In Forsch. 
Geb. Agr. Phys., 13 : 45. 1890. 
9. Franke, . In Ann. Agron., Tome 2. 

10. Fred, E. B., and Hart, E. B. Comparative effect of phosphates and sul- 

phates on soil bacteria. Wis. Agr. Expt. Sta. Research Bui. 35. 

11. Gardner, Frank D. Fertility of soils as affected by manures. U. S. Dept. 

Agr., Bur. Soils Bui. 48, p. 54. 1908. 

12. Hall, A. D., and Miller, N. H. In Proc. Roy. Soc, ser. B 77, p. 1. 1905. 

13. Hellreigel, , and Wilfarth, . In Ann. Agron., Tome 15. 

14. Hopkins, C. G. Soil fertility and permanent agriculture, p. 195. Ginn & 

Co., Boston. 1910. 

15. Kaserer, Hermann, liber die Oxydation des Wasserstoffes und des 

Methans durch Mikroorganismen. In Centr. Bakt. 15: 573 (1905); 16: 
681 (1906). 

16. Knudson, Lewis. Influence of certain carbohydrates on green plants. N. 

Y. Cornell Agr. Expt. Sta. Memoir 9. 1916. 

17. Laurent, M. J. Recherches sur la nutrition carbonee des plantes vertes a 

l'aide de matieres organiques. In Rev. Bot, 16: 14-48, 96-117. 1904. 

18. Lefevre, Jules. Sur la developpement des plantes vertes a la lumiere en 

l'absence complete de gas carbonique dans un sol artificial des amides. In 
Comptes Rendus, 141 : 211-213, 664, 665. 1905. 

19. . Ibid., p. 834, 835. 

20. . Ibid., p. 1035, 1036. 

21. Lemmermann, O., Aso, K., Fischer, H., and Fresenius, L. Unter- 

suchung iiber die Zerzetzung der Kohlenstoff verbindungen verscheidener 
organischen Substanzen im Boden spezielle unter dem Einfluss der Kalk. 
In Landw. Jahrb., 41: 216-257. 1911. 

22. . Op. cit, p. 244. 

23. Lohnis, F. Boden Bakterien und Boden Fruchtbarkeit. Berlin. 

24. Mitscherlich, E. A. Ein Beitrag zur Kohlensaurediingung. In Landw. 

Jahrb., 39 : 157-166. 1910. 

25. Molliard, M. Culture pure des plantes vertes dans une atmosphere con- 

finee en presence des matieres organiques. In Comptes Rendus, 141 : 
389-391- 1905. 

26. . L'humus est il une source de carbon pour plantes vertes superieuse? 

In Comptes Rendus, 154: 291-294. 1912. 

27. Mooers, C. A., Hampton, H. H., and Hunter, W. K. The effect of liming 

and of green manuring on the soil's content of nitrogen and humus. 
Tenn. Agr. Expt. Sta. Bui. 96, parts 2 and 3. 

28. Nabokish, A. J., and Lebendeff, A. F. liber die Oxydation des Wasser- 

stoffes durch Bakterien. In Centr. Bakt., Abt. 2, 17: 350. 1907. 

29. Pfeffer, Wilhelm F. P. Physiology of plants, v. 1, p. 171. 1899. 

30. Pollacii, G. Nuove recherches sul l'assimilazione del carbonic In Bui. 

Soc. Bot. Ital. (1911-12), p. 208-211. 1912. 

31. Potter, R. S., and Snyder, R. S. Carbon and nitrogen changes in soil 

variously treated with ammonium sulphate and sodium nitrate. In Soil 
Science, v. 1, no. 1, p. 76-94. 1916. 

32. Quarrie, George. The application of C0 2 gas to the soil. In Sci. Amer. 

Supp., p. 339- I9M. 



3C2 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



33. Ravin, . Nutrition carbonee des phanerogames a l'aide de quelques 

acids organiques et de leur sels potassiques. In Comptes Rendus, 154: 
1100-1103. 1912. 

34. Reichardt, , and Blumtritt, . In Jour. Prakt. Chem., 98: 476. 

1866. 

35. Reisset. . In Comptes Rendus, 88, 1001 ; 90: 1144. 

36. Russell. E. J. Oxidation in soils and its connection with fertility. In 

Jour. Agr. Sci., 1 : 261-279. 1905. 

37. Schlossing, , and Laurent, . In Ann. Inst. Pasteur, tome 6. 

38. Schumacher, Wilhelm. Ernahrung des Pflanze, p. 76. Berlin, 1864. 

39. Schreiner, O., Shorey, E. C, Sullivan, M. X., and Skinner, J. J. A 

beneficial organic constituent of soils : Creatinine. U. S. Dept. Agr., 
Bur. Soils Bui. 83. 

40. Sutton, Francis. Volumetric analysis, 8th ed., p. 496. 1900. 

41. Van Suchtelen, F. H. H. Uber die Messung der Lebensthatigkeit der 

aerobischen Bakterien im Boden durch die Kohlensauerproduktion. In 
Centr. Bakt., Abt. 2, Bd. 28, S. 45. 1910. 

42. Vox Dobeneck, Arnold F. Untersuchungen iiber das Absorptionsver- 

mogen und die Hygroskopizitat der Bodenkonstituenten. In Forsch. 
Geb. Agr. Phys., 15: 201. 1892. 

43. Walker, G. W. Minn. Agr. Expt. Sta. Tech. Bui. 128, p. 179. 

44. Wollney, E. Die Zerzetzung der Organischenstoffe, p. 405. 1897. 



CROSS-POLLINATION OF SUGAR CANE. 1 

H. B. Cowgill. 

Sugar cane has been propagated from seed and the seedlings se- 
lected for the purpose of originating new varieties since 1887. This 
was begun in Java and in Barbadoes at about the same time, and it 
has since been taken up in nearly all the cane-producing countries of 
the w orld. Originally no record was made of the parentage of the 
seedlings, and in many cases not even the name of the seed parent was 
kept. Some very good varieties were originated by this method. 

For commercial purposes cane is propagated asexually by cuttings. 
When il 19 propagated from seed the variation in the resulting gen- 
eration, even from a single parent v.ariety, is considerable. It is pre- 
sumed that sonic, if not all, of the varieties are more or less hetero- 
lt seems nevertheless desirable, in many cases, to make 

controlled crossei in order to combine such characters as vigor and 
tance of certain varieties with good qualities of other 

varieties. 

1 iitnlmtion from the I'orto Kiro Insular Experiment Station. Received 

f<»r jmhliration July 23, 191K. 



COWGILL : CROSS-POLLINATION OF SUGAR CANE, 



303 



Methods of Crossing. 
It would, of course, be desirable to eliminate all possibility of self- 
pollination. Attempts to emasculate the florets have been made, and 
a few seedlings have been produced in Barbadoes in that way ; though, 
according to Bovell, the number of seedlings produced in any single 
season has been small. The work is very tedious, for the reason that 
the florets are small and the panicle is brittle. The latter is also pro- 
duced at 10 to 15 feet from the ground, so that it is necessary to do 
the hybridizing on a scaffold and sometimes the wind makes the work 
very difficult. 

Kobus (4), 2 in Java, planted a pollen-sterile variety on the leeward 
side of a pollen-fertile variety which flowered at the same time. 
Seeds of the former, when planted, grew and developed into canes 
which had characteristics of both parent varieties. 

Another method reported by Bovell (1) to be employed in Barba- 
does is to plant two varieties which flower at the same time in alter- 
nate stools, called the "checkerboard system," for the purpose of facili- 
tating natural cross-pollination. It is of course impossible to form 
any conclusion as to the extent to which crossing takes place with this 
method, unless the type of seedlings produced by each variety when 
growing separately is known. 

Two additional methods are described by Wilbrink and Ledeboer 
(6). By the first method the tassels of the variety to be used as the 
male parent are cut off and tied in position with the one to be used as 
a seed parent. For protection against undesired pollen a screen is 
provided, having an opening on the leeward side for the entrance of 
the tassels. By the second method the pollen of the desired variety 
is collected and carried to the one to be used as the female parent. 
This later method is also one which was suggested by D'Albuquerque 
(3). It is reported that the pollen adheres in masses, and also soon 
deteriorates, so that no very satisfactory results were obtained. 

Methods Employed at the Insular Experiment Station. 

Crossing has been practiced at the Insular Experiment Station of 
Porto Rico for four years. The method here described was found 
to be more suitable, for the reason that with its use a fairly large 
number of seedlings can be produced. The work has not yet pro- 
gressed far enough to report results of the crossing, in respect to the 
quality of varieties produced. It has been possible, however, to study 
to some extent the populations of seedlings originating from different 
2 Figures in parentheses refer to " Literature cited," p. 306. 



304 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

parentages, as to inheritance of characters in first-generation seed- 
lings. 

Bags made of cheese-cloth are held extended by heavy wire rings 
sewed into them. The bags when completed are 48 inches long and 
18 inches in diameter. The rings are placed one at the top and the 
other 16 inches from the bottom, so that a skirt of 16 inches is left 
to be drawn in and tied around the stems of the panicles. 

The bags are supported over the panicles by means of bamboo poles 
set in the ground. The poles have a crossbar at the top which is 
fastened to them by being wedged into notches cut into the second 
internode from the top, and the bags are tied to this crossbar. The 
poles are set on the windward side of the stools just before the pan- 
icles " shoot ; " when this occurs, a bag is immediately suspended over 
each panicle and tied around its stem, so that it is protected from all 
undesired pollen before any of the florets open. 

The cane blossom is hermaphrodite, but it has been found that cer- 
tain varieties are almost completely pollen-sterile, or at least self- 
sterile. This makes it possible to pollinate them with another variety, 
with the assurance that nearly all the seedlings will be offspring of 
two known parent varieties, a few usually also being produced as the 
result of the self-pollination of the mother parent. 

The pollinating is done by placing panicles of the desired variety 
into the bag, in such a position that their pollen will be shed or carried 
by the wind to the florets of the other variety as they open. One or 
two panicles are used at a time, and they are allowed to remain in the 
bag two or three days, being renewed as often as necessary while the 
florets arc opening. It has been found of advantage to cut the pan- 
idea with stems 4 to 6 feet long, and to place their lower ends in a 
joint of bamboo filled with water, by which they can be kept fresh 
two or three clays. 

Results Accomplishkd. 

rp to the present time, results can only be expressed in terms of 
the number oi seedlings produced and the extent to which the char- 
of the varieties arc combined. The method above described 
rst tried in 191 5-16. Ten crosses were attempted, of a single 
Combination, and all but two produced seedlings, a majority of which, 
•aIicu mature, ifiowed characteristics ( ,f both parents. Tn all, about 
M-cdlin^ were produced, one panicle alone giving over 1,000 
seedlings (2). 

In the following winter of 1916-1917, thirty crosses, comprising 
nine differ* 111 combinations, were attempted, and nineteen of them, 



COWGILL : CROSS-POLLINATION OF SUGAR CANE. 



3°5 



comprising six combinations, were successful. From one combina- 
tion 1,309 seedlings were obtained, and in all 2,589 seedlings were 
produced. The work was all done by one man and a helper, includ- 
ing the making of the bags. 

In 1917-1918 it was impossible to secure the services of a competent 
man to perform the crossing until late in the season, and the seed of 
all varieties was also much less viable than in the preceding year. 
Thirty crosses were attempted, comprising nine combinations. Fif- 
teen of these were successful and 1,794 seedlings were produced, 857 
of which were from one combination and 735 from another. 

Judging from the small proportion of the seedlings out of the large 
number propagated by the old method that are of sufficient value to 
become widely cultivated, it appears that a large number of first-gen- 
eration seedlings is essential. Considered from the point of view of 
Mendelian inheritance, if many factors are involved, which is prob- 
ably the case, the chance of getting a desired combination of charac- 
ters is very remote when only a few seedlings are grown. 

Effect of the Crossing. 

In 1915-1916 the variety used as a pollinator was a dark-colored 
cane, while the seed parent was medium light. This made it possible 
to trace the color of the male parent in the offspring. Some other 
characters could also be traced in the seedlings in the same way. In 
the following year this cross was again made, and the same general 
effects were observed, many of the same types being again recog- 
nized (2). 

In the year 1916-1917, some of the parent varieties of groups of 
seedlings showed fewer differences than was the case with the varie- 
ties combined the year before, consequently it was less easy to see 
the effect of the crossing in the seedlings. In all cases but one, how- 
ever, some of the groups showed distinguishing characteristics of both 
parent varieties. 

The disadvantage in this method, in not being able to eliminate 
all possibility of self-pollination, ought not to be overlooked. On ac- 
count of the chance of some selfing, it has been the practice to esti- 
mate the value of a cross from the entire group of seedlings produced, 
always making allowance for probable self-pollination. 

Self-Sterility. 

At least two of the old standard varieties are nearly pollen-sterile 
here. We have never succeeded in producing more than one to five 



306 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



seedlings from single flats of several hundred seeds planted, while if 
these same varieties are pollinated by any of several seedling varie- 
ties good germination follows. Lewton-Brain (5) in Barbadoes ex- 
amined the florets of about fifty varieties and found that some bore 
pollen nearly all of which was large, well-shaped and full of dark 
granules, while with some the pollen was smaller, more or less irregu- 
lar in form, and without granular matter. A third class of varieties 
had an intermediate amount of normal, well-developed pollen. 

Wilbrink and Ledeboer (6) describe a method of testing the pollen 
with iodine, to determine its viabilty. If the pollen grain contains 
starch it was believed to be normal. We have not, however, found 
this test to be absolutely reliable. 

Conclusions. 

From the work reviewed in the foregoing paper the following con- 
clusions are possible : 

1. Sugar cane can be cross-pollinated and protected from outside 
pollen, and by this process a considerable number of seedlings can be 
produced. 

2. Characters of the parent varieties are combined in the seedling 

by this process. 3 

It should be expected that the desirable combinations could be perpetuated 
in hybrid condition because of the asexual method of propagation, a rather 
unusual advantage among our field crop plants. (Editorial note by L. H. 

Smith.) 

Literature Cited. 

1. Bovell, J. R. Report of sugar cane experiments. In Rpt. Barbadoes Dept. 

Agr., p. 15-16. 1914. 

2. ( "V m l, H. B. Studies in inheritance in suger cane. In Jour. Dept. Agr. 

Porto Rico, vol. 2, no. i, p. 33-41. 1918. 



3 ])' \i m <ji erque, J. P. Note on the artificial cross-fertilization in the sugar 

cane. In West Indian Bulletin, vol. 1, no. 2, p. 182-184. 1900. 

4- Kobi J. I). De zaadplanten de kruising van Cheribonriet met de Englisch- 
[ndifhe Varietiel Chunnee. In Mededcelingen vat bet Proefstation voor 
de Java-Snikerindustrie, ser. Ill, nos. 1, 12, 21, 33. 

.'• Leu on BraxXj I.. Hybridization of sugar cane. In West Indian Bulletin, 

vol. 4, no. 1, p. 63-72. 1903. 
6. WlLTON X and LXDEBOER, F. De geslachtelijke voortplanting bij bet Suik- 
h Mc-dcdcclin^'cn van bet Proefstation voor de Java-Suikcrin- 
dtistric, no. (>. i<ji 1 . 




AGRONOMIC AFFAIRS. 



307 



AGRONOMIC AFFAIRS. 



ANNUAL MEETING IN BALTIMORE. 



The eleventh annual meeting of the American Society of Agronomy 
will be held in Baltimore, Md., November 11 and 12. The call for 
papers for the program has been sent out by the Secretary. All who 
expect to present material at this meeting are urged to send titles 
at once to Lyman Carrier, Department of Agriculture, Washington, 



The editor's attention has been called to an error which occurred 
in the article by Dr. George F. Freeman in the January issue of the 
Journal. In Table 2, page 25. the words "hard" and "soft" are 
transposed throughout, so that the data are made to show just the 
opposite tendency from that actually indicated. All readers are re- 
quested to change the column headings in this table as here indicated. 



The membership reported in the September issue of the Journal 
was 652. Since that time 9 new members have been added and 10 
have resigned, making the present membership 651. Names and ad- 
dresses of new members, names of members resigned, and changes 
of address which have been reported follow. 



Birchard. J. F.. Cor. Magnus Ave. and Main St.. Winnipeg, Manitoba. 
Blackwell, C. P., Clemson College, S. C. 
Damon, S. C, Agr. Expt. Sta., Kingston. R. I. 
Elliott, B. S. A.. School of Agr., Olds, Alta., Canada. 
Fergus, E. X., Agr. Expt. Sta.. La Fayette, Ind. 
Grisdale. F. S.. School of Agr., Vermilion, Alta., Canada. 
Moxteaguda, Heriberto, Quinta de los Molinos, Habana. Cuba. 
Stephex. YV. J., School of Agr.. Claresholm. Alta., Canada. 
Van Schaik, K. L., Pretoria. Transvaal, South Africa. 



D. C. 



ERROR IN THE JANUARY NUMBER. 



MEMBERSHIP CHANGES. 



Xew Members. 



Davissox, B. S., 
Deeter. E. B.. 
Gray, Wm. F.. 
Hitchcock, E. B. 



Members Resigxed. 
Lyxess, W. E., 

X'ORTHROP, ROBT. S., 

Sherbakoff, C. D.. 



Sleeth, E. C, 
Young, Horace J., 
Zahxley, J. W. 



3oS 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Changes of Address. 

Alexander, L. L., State Normal School, Springfield, Mo. 

Bauer. F. C. 218 N. Lake St., Madison, Wis. 

Beavers. J. C, Guilford College, N. C. 

Bugby, M. O., Kingsville, Ohio. 

Farrell, F. D., Agr. Expt. Sta., Manhattan, Kans. 

Frear, D. W., Crop Estimates, U. S. Dept. Agr., Washington, D. C. 

Freeman, Geo. F., Societe Stiltanienne d'Agriculture, Cairo, Egypt. 

Gerxert, W. B., Farm Bureau, Paris, 111. 

Goddard. L. H., Washington Ave., Washington C. H., Ohio. 

Jexsex, L. X., Box 308. Big Springs, Tex. 

Luxd. Viggo, Tystoftu Expt. Sta., Tjaereby, Denmark. 

Mexzies, E. C, Aylesburg, Sask., Canada. 

Mixer, Sterling L., 1029 Sixth St., Greeley, Colo. 

Petrv. E. J., 210 South Ingalls St., Ann Arbor, Mich. 

Southwick, Benj. G., Mass. Agr. College, Amherst, Mass. 

Stadler, L. J., 410 South Maple St., St. Louis, Mo. 

Thatcher, Lloyd E., Ohio Agr. Expt. Sta., Wooster, Ohio. 

Tillman, B. W., Univ. of Mo., Columbia, Mo. 

Walster, H. L., Soils Bldg., College of Agriculture, Madison, Wis. 



ROLL OF HONOR. 

The Society's roll of honor of men in military service now contains 
the names of 47 men. No doubt there are many others whose names 
have not been reported to the officers of the Society. The editor will 
appreciate information regarding men in the service of their country, 
both items of news and lists of names of men from various institu- 
tons who are now engaged in war activities. The names of those 
who have been reported to the editor follow. 



Ai.i'.j IT, A. R., 

Andrews, Myron E., 
Buss, S. W., 
Brockson, W. I., 
Bruce, O. C, 
Brunson, A. M., 
Burnett, Grover, 
c "ati , Hi NRY k'., 
Chapman, James E., 
Guilds, R. R„ 
I Patrick, E. P., 
H Werff, II. A., 
Dickenson, R. W., 
Downs, E. E., 
Ellison, A. D., 
Freeman, Ray, 



Gentle, G. E., 
Gilbert, M. B., 
Graham, E. E., 
Gray, Samuel D., 
Hanson, Lewis P., 

I lol l. AM), B. B., 

1 1 1 i»i i. sox, R. R., 

Jl.XSKN, O. F., 

K 1 N worth y, Chester, 
Kimi. P. H„ 

Macfarlan I , Wallac I 

MOOM \w, I j AOY, 

Newton, Robert, 
Palmer, ii. Wayne, 

PlI MI ISEL, R. L., 

Pi El rOTON, Jami s A., 



QUIGLEY, J. V., 

Ratcliffe, Geo. T., 
Raymond, L. C, 
Richards, Phil E., 
Schneideriian, F. J. 
Sciioonover, W. R., 
Scott, Herscbel, 
Smith, J. P., 
Spencer, E. L., 
Stanley, C. W., 
Starr, S. ii., 

Tahor, Paul, 
Towi.e, K'. S., 

Ware, J. ()., 

Wi s 1 IROOK, E. C. 



AGRONOMIC AFFAIRS. 



309 



NOTES AND NEWS. 

E. C. Chilcott, C. S. Scofield, and T. H. Kearney of the Department 
of Agriculture are now in Algeria, Tunis, and Morocco, where they 
are investigating the possibilities of increasing the agricultural output 
of these French colonies. The trip is being made at the request of the 
French high commission to the United States. 

G. I. Christie, director of extension in Indiana and for the past sev- 
eral months assistant to the secretary of agriculture, has been named 
as assistant secretary, succeeding R. A. Pearson, resigned to resume 
his duties as president of Iowa State College. 

G. H. Collings has been appointed assistant professor of agronomy 
and assistant agronomist at the Clemson (S. C.) college and station. 

George F. Freeman, plant breeder of the Arizona station, has re- 
signed to take up cotton-breeding and cultural work for the Egyptian 
Government in the valley of the Nile. 

Ben C. Helmick, formerly assistant professor of agronomy in the 
Minnesota college of agriculture, is now instructor in agronomy in 
the Connecticut college and associate agronomist of the Storrs station. 

C. G. Hopkins has been granted a year's leave of absence from the 
Illinois college and station to head the agricultural section of the Red 
Cross commission to Greece. He will study soil conditions, particu- 
larly with a view to the quick and permanent increase in food produc- 
tion. He will be assisted by George Bouyoucos of the Michigan sta- 
tion, the son of a Greek farmer, who has had fifteen years training 
and experience in this country. 

W. L. Hutchinson, for the past several years professor of agron- 
omy at Clemson College, has resigned and has been succeeded by C. 
P. Blackwell, who will also be agronomist of the South Carolina 
station. 

J. S. Jones, formerly director and chemist of the Idaho station, has 
resigned to take charge of the operating laboratory of one of the Gov- 
ernment's nitrate plants under the ordnance division of the War De- 
partment. 

F. M. Rast, jr., formerly assistant professor of soils and fertilizers 
in the University of Florida, has been appointed assistant professor 
of agronomy in the Delaware college, succeeding M. L. Nichols, who 
is now in charge of extension work in farm engineering in Virginia. 

Benjamin G. Southwick, formerly of the Connecticut college, is now 
demonstrator in farm management in the Massachusetts college. 

J. L. Staden and C. C. Hearne have been appointed assistants in 
farm crops in the University of Missouri. 



3IO JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

L. E. Thatcher has resigned as instructor in farm crops in Ohio 
State University and is now assistant agronomist in charge of plant- 
breeding work at the Ohio station. 

B. W. Tillman, formerly of the soil survey force of the Federal 
Department of Agriculture, has resigned to become extension as- 
sistant professor of soils in Missouri. 

H. L. Walster, who has been pursuing graduate study in plant 
physiology and ecology at the University of Chicago, has returned to 
his work at the University of Wisconsin, having received the Ph.D. 
degree at the August convocation. He will devote most of his time 
in Wisconsin to plant nutrition studies. 

A committee to obtain information on food production conditions 
in Great Britain, France, and Italy, with a view to making these con- 
ditions known to agricultural leaders and to farmers generally in this 
country and to enable tfs to render aid more effectively to these coun- 
tries, arrived in England September i. The committee consists of 
Dr. W. O. Thompson, president of Ohio State University, chairman ; 
Carl Vrooman, assistant secretary of agriculture ; R. A. Pearson, 
president of Iowa State College ; T. F. Hunt, director of the Califor- 
nia station ; D. R. Coker, farmer and member of the national agri- 
cultural advisory committee ; W. A. Taylor, chief of the Bureau of 
Plant Industry ; G. M. Rommel, of the Bureau of Animal Industry ; 
and George R. Argo and John F. Wilmeth, of the Bureau of Markets. 

Probably for the first time in history, the Federal Government has 
been making direct loans to farmers to finance fall seeding of wheat 
and rye. These loans are being made in Montana, North Dakota, 
Kansas, Oklahoma, and Texas, where crop failures during the past 
two years have been general over wide areas. Only those who can 
obtain fund- in no other way are being financed. The work is under 
the joint supervision of the Treasury and Agricultural departments, 
the applications being approved by the Department of Agriculture 
and the loans completed by the Treasury Department, through the 
Federal land banks in these districts. G. I. Christie, assistant to the 
lecretary of agriculture, lias been in charge of the. work in the north- 
ern district and I.. M. Kstabrook, chief of the bureau of crop csti- 

raates, hai supervised tlx- making of loans in the southwest. C. W. 

Warbnrton and II. V Yinall have been chief assistants to Messrs. 
( hristie and Kstabrook. re pectivcly. The loans are being made from 
the I 'resident's war emergency fund, $5,000,000 having been set aside 
for that purpose. 'I 1h- portion of the fund not loaned this fall will 
In n ''I for financing seeding of spring wheat. 



AGRONOMIC AFFAIRS. 



Third Western Agronomic Conference. 

Agronomic workers in the eleven western states met in the third western 
agronomic conference at Corvallis, Oreg., July 23, 24, and 25, 1918. The meet- 
ings were attended by more than forty agronomic workers representing the 
various State experiment stations, agricultural colleges, and extension depart- 
ments, the United States Department of Agriculture, and the British Columbia 
Department of Agriculture. 

Roland McKee of the Office of Forage Crop Investigations, U. S. Depart- 
ment of Agriculture, stationed at Chico, Cal., discussed experimental methods 
for establishing forage standards. He emphasized the importance of basing 
forage yields on the oven-dry basis because of varying rates of loss of moisture 
after cutting with different varieties, and also with light and heavy yields of 
the same variety. H. A. Schoth of the same office, stationed at Corvallis, 
Oreg., discussed production problems in connection with forage-crop experi- 
ments. 

Experimental work in the prevention of smut explosions and in grain clean- 
ing in the smut-infested areas of the Northwest, as well as means of collecting 
smut spores to prevent soil infection, was presented by C. C. Ruth, of the Port- 
land grain supervision office, Bureau of Markets. H. P. Barss, professor of 
botany and plant pathology of the Oregon college, discussed the smut and rust 
control work under way in the United States and the splendid results accom- 
plished by the pooling of interests of the plant pathologists throughout the 
country. 

R. L. Stewart of the New Mexico college presented a very interesting paper 
on the utilization of soaproot and sotol as forage for range cattle. He stated 
that thousands of head of cattle had been carried through the period of forage 
shortage by feeding these plants finely chopped. Range management in 
\Vyoming was discussed by A. F. Vass, agronomist of the Wyoming station. 
He showed the direct relation between dry }^ears and loss of stock and also 
the relation between prices and the rise and decline of the stock population of 
the State. 

O. E. Barbee of the Washington station presented a season's data on the 
influence of date of seeding on winter wheat. Row versus plat plantings for 
varietal test's of cereals were discussed by D. E. Stephens, superintendent of 
the Moro, Oreg., substation. Reliable results have been obtained at this sta- 
tion from row tests. 

J. A. Clark, agronomist in charge of western wheat investigations for the 
U. S. Department of Agriculture, presented the scheme of classification of 
wheat varieties worked out by C. R. Ball and himself, and then illustrated the 
scheme by taking the visiting agronomists to the wheat nursery of the Oregon 
station, where the various commercial varieties of wheat of the United States 
were being grown. 

Soil problems in the Palouse district, particularly with reference to humus 
and nitrogen, were discussed by F. J. Sievers of the Washington college. Ex- 
perimental work with alkali soils in Utah was presented by B. W. Pittman, 
and soil survey problems in Oregon and Idaho were discussed by C. V. Ruzek 
and E. B. Hitchcock, respectively. The follow-up work after the mapping and 
physical classification were completed was stated to be of maximum importance. 



3** 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



W. L. Powers of the Oregon college presented a paper on some phases of 
experimental work in farm management, and George Stewart of the Utah 
college talked of irrigation investigations being conducted in that State. The 
results of an extensive survey of problems of the field-pea growers of Idaho 
were presented by H. W. Hulbert. 

A round-table discussion on the bulk handling of grain in the Pacific North- 
west was led by A. L. Rush of the Bureau of Markets and G. R. Hyslop of 
the Oregon college and was participated in by many of those present. The 
discussion of factors affecting quality of wheat was introduced by J. W. Gil- 
more of the University of California. It was agreed by those present that 
quality of wheat was dependent on both soil and climatic conditions. 

H. D. Scudder, professor of farm management in the Oregon college, pre- 
sented plans for land colonization in Oregon, particularly with reference to a 
model farm colonization unit. 

Farm crops work offered in the various western colleges was presented in 
tabular form and discussed by E. G. Schafer of the Washington college, par- 
ticularly with regard to its relation to other required work. Agronomic ex- 
tension work was discussed by Leonard Hegnauer and R. J. Leth, extension 
specialists in agronomy in Washington and Idaho, respectively. The impor- 
tance of extension correspondence was particularly emphasized. 

The invitation of Professor Gilmore to hold the 1919 conference in Berkeley 
was accepted, and the date of the meeting was set for early June. Professor 
Gilmore was elected chairman of the 1919 conference, and named Roland Mc- 
Kee and G. R. Hyslop as additional members of the committee on arrange- 
ments. As some of the agronomic meetings in the Middle West have now 
been abandoned, it is hoped that agronomists from that section will join with 
western agronomists in the 1919 conference at Berkeley. 

G. R. Hyslop, Secretary. 




JOURNAL 

OF THE 

American Society of Agronomy 



Vol. io. December, 1918 No. 9. 



INFLUENCE OF HIGHER PLANTS ON BACTERIAL ACTIVITIES 

IN SOILS. 1 

T. Lyttleton Lyon. 

In considering the relation of plants to soils attention has in general 
been more concerned with the effect of the soil on the plant than with 
the influence which the plant may exert on the soil. When, however, 
the -latter has received consideration it has been mainly from the 
standpoint of the quantity of fertility removed. In more recent 
years investigations have been conducted to ascertain the effects of 
green-manuring crops or of systems of crop rotations on nitrogen 
transformations in soils. 

The subject to which I wish to call your attention is the immediate 
influence of the growing plant on certain bacterial processes in the 
soil, a subject which has a bearing on the practical problems of crop 
production and which is now in that interesting stage of suggestion 
that is so alluring to the investigator. 

The literature of soil investigation is not rich in indications that the 
bacterial flora and its activities are influenced by growing plants, but 
some such suggestions may be found, and they give a reasonable basis 
for encouraging further investigations. I shall not attempt an ex- 
haustive review of this literature, but shall mention some of the work- 
that appears most significant. I find only a few investigations that 
attempt any correlation between the growing plant and the bacterial 
flora. One is a piece of work by Hoffmann in which he counted the 
number of bacteria in soil immediately adjacent to plant roots and 

1 Presidential address before the eleventh annual meeting of the American 
Society of Agronomy, January 6, 1919. Read by C. E. Leighty in the ab- 
sence of President Lyon. 

313 



314 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



at some distance away. He examined a large number of plants of 
different genera and species and found almost uniformly a denser 
flora near the roots. He is also supported by Stoklasa, who found 
under different crops marked differences in the germ content of the 
soil. It will, of course, require a very considerable number of tests 
to establish the certainty that roots of growing plants improve the 
surrounding soil for the development of bacteria and that certain 
higher plants are more effective than others in this respect. It is men- 
tioned here, however, because the possibility is suggested rather 
strongly. 

A class of microorganisms whose activities, there is reason to be- 
lieve, are influenced by growing crops is that portion of the flora 
which reduces nitrates to less highly oxidized forms of nitrogen. An 
unaccountable disappearance of nitrates from soils on which crops 
were growing has been noted by several investigators. A number 
of explanations have been proposed. Deherain attributes it to the 
drying out of the soil by the growing plants during the season most 
favorable for nitrification. Warington thought that it might be due 
to denitrification with loss of nitrogen into the air and possibly to loss 
of nitrogen from the plant. Russell merely states that it indicates a 
diminished production of nitrates and Leather is inclined to accept 
the explanation which will be given later. 

A typical example of this disappearance of nitrates may be found 
in the lysimeter experiments of Deherain in France, Leather in India, 
and of the Cornell experiment station. In all of these experiments 
certain plants grown in lysimeters from which both the plants and 
drainage water were analyzed gave as a balance at the end of the 
experiment less nitrogen in the crop plus the drainage water of the 
planted soil than in the drainage water alone of the soil on which 
no plants grew. No analyses having been made of the soil it is 
uncertain what finally became of the nitrogen that failed to appear 
either in the crop or in the drainage water. 

It i- evident that the problem must he attacked in a different way 
and to do this plant- were grown in nutrient solutions which were 
kept Merile through the entire period of the growth of the plants, 
amounting in some caw- to nearly a year. In these solutions con- 
tained in 12-liter flasks plants were grown to maturity and maize 
reached a height of (> feet. The solution being sterile when the plant 

■\ har\( ted, it could he used as a medium in which to conduct bac- 
terial transformations of nitrogen with pure cultures. To study the 
iU di appearance <>\ nitrate nitrogen the solution in which the 
plant had grown had added to it a definite quantity of a nitrate salt 



lyon: influence of plants ox soil bacteria. 



315 



and, after inoculation with a nitrate-reducing organism, it was in- 
cubated and the quantity of nitrate that disappeared in a certain time 
was ascertained by analysis. This was compared with a similar 
nutrient solution in which no plant had grown. Without taking the 
time that would be consumed in presenting the figures involved in 
this and other experimental work described in this paper I may say 
that there appeared to be a markedly more rapid reduction of nitrates 
in the solution in which the plant had grown than in the other, the 
experiment being repeated many times. 

The composition of the nutrient solution in which the plant had 
grown and the one in which it had not was, of course, different on ac- 
count of the removal of certain substances by the plant. To make the 
two as nearly similar as possible 10 c.c. of the solution in which the 
plant had grown was added to 90 c.c. of another medium. The same 
was done with the solution in which no plant had grown and under 
these conditions the solution in which the plant grew increased the 
rate of nitrate reduction more than did the solution in which no plant 
grew. Analyses of the solutions after incubation showed that part of 
the nitrates had been converted into organic forms. 

Infusions of macerated plant roots were made and reduction of 
nitrates tested in solutions to which these were added. It was found 
that such infusions increased nitrate reduction and that the more 
infusion added the more rapid was the disappearance of nitrate. 
Mannite was used in the incubated solutions as a source of energy 
for the bacteria and it was found that within certain limits an in- 
crease in the quantity of mannite served to hasten nitrate formation. 
The organic matter available to the reducing organisms is doubtless 
a factor in determining their activity. 

In solutions in which plants grew as well as in media to which in- 
fusion of macerated roots was added it was attempted to ascertain 
whether nitrate reduction went on in the presence of an antiseptic, 
phenol being used. Such reduction did occur while it did not do so 
in check solutions, but it was slight and the evidence indicates that the 
effect of growing plants on nitrate reduction is due only in part to 
enzymotic action. Reducing enzymes were generally found, however, 
in plant solutions and in root infusions. Oxidates were sometimes 
shown to be present in plants growing in agar, timothy roots always 
giving reaction for these. Reactions for peroxidases were always ob- 
tained in agar near the roots of all plants tested. Boiling the solu- 
tions before inoculation lessened nitrate reduction, thus indicating 
that enzymes or some similar substances played a part in the process. 



3l6 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

The experiments that have been described indicate two possible 
ways in which higher plants may influence bacterial activities in soils. 
One of these is through the result of plant growth on the composition 
of the soil solution. Experiments by a number of investigators show 
that the relative quantities of certain anions and cations in solution 
may influence the rate of the nitrifying process. The presence or 
absence of organic matter has also been shown to be a factor. 

The experiments that have been mentioned indicate that both of 
these conditions may influence reduction of nitrates. The growing 
plant may be expected to have an influence on the composition of the 
soil solution. From the time that the plant begins to grow until it 
reaches the stage of full bloom it is absorbing nutrients from the soil 
solution with slight if any return to the soil. During this period there 
is doubtless a decrease in the concentration of the soil solution, as ab- 
sorption apparently proceeds faster than solution. After the stage 
of full bloom a marked diminution in the absorption of the solutes 
begins. A change in the composition of the soil solution may be 
looked for during the later stages of growth of the plant. It is 
during the middle and later stages of growth that the reduction of 
nitrates appears to be most marked. Plants of different kinds also 
differ in respect to the quantities and rates of nutrients absorbed and 
may thus be expected to exert different effects on the rate of reduc- 
tion of nitrates. 

Another possibility is that the organic matter in the soil solution is 
more or less influenced by plant growth and that this organic matter 
affects the activities of the bacteria that bring about the transforma- 
tions of nitrogen. The organic matter in solution constitutes a very 
small part of the total organic matter of the soil. Being in solution it 
is in a condition to affect bacterial activity. In the experiments pre- 
viously mentioned there was found in the solutions in which plants 
had grown a half percent as much organic nitrogen as was found in 
the plant at maturity, altho these solutions were originally composed 
Only of Inorganic substances. This organic matter given off by the 

plan! root! may conceivably be a factor in the depressing effect on 
the nitrate content of soils. 

In trying to determine the action of any plant on transformations 
of nitrogen little can he learned by making incubation tests of the 
loil placed in flasks. The aeration and partial drying which the soil 
lindergoe in the process is likely to nullify any effect which the crop 
may have had. [til well known thai either complete or partial drying 



LYON: INFLUENCE OF PLANTS ON SOIL BACTERIA. 3I7 

of a soil results in materially increasing its solubility. Any effect that 
the plant may have exerted on the composition of the soil solution 
would therefore be changed and nitrogen transformations may be very 
different from what they would have been in the untouched soil. Even 
the operation of plowing causes a change in the rate of nitrate forma- 
tion, as has been shown by Brown and Maclntire. The best way to 
determine the ultimate effect of a crop on nitrate formation is to re- 
move the crop and allow the soil to incubate without disturbing it, at 
the same time maintaining an optimum moisture content. 

Another bacterial process in soils which appears to be influenced by 
some higher plants is the production of carbon dioxide. Under cer- 
tain conditions the carbon-dioxide content of the air of soil on which 
plants had matured was lower than that of soil on which no plants 
had grown, altho the opposite was the case during the time the 
plants were making their greatest growth. Both nitrate production 
and carbon-dioxide formation are associated with the decomposition 
of organic matter and it would thus appear that this process is in a 
measure at least controlled by crop growth. 

Xot only does there appear to be a depression of nitrate production 
by certain higher plants, but other of these plants seem to have a 
stimulating influence on the formation of nitrates. This, however, 
appears to be exerted only during the early stages of growth. There 
is not so much evidence regarding this property of plants as there is 
regarding their depressive action, but there is some indirect exper- 
imental indication of its occurrence. 

Fraps found that 50 to 100 percent more nitrogen was removed 
from the soil by maize plants in the first nine weeks of their growth 
than was apparently transformed from organic compounds into am- 
monium and nitrate salts during the same time. 

The figures given by Stewart and Greaves in a study of nitrates in 
irrigated soils planted to maize, potatoes, and alfalfa, and also on land 
fallowed during a period of three years and planted to oats one year, 
show nitrates to be higher under maize at certain stages in the growth 
of the crop than in fallow land. The same was true of potatoes in 
their later experiments. 

Results reported by Jensen showed that soil planted to maize con- 
tained more nitrates during the first part of July than did fallow land. 

Bower determined nitrates in soil of unplanted plots both cultivated 
and uncultivated, and in maize plots cultivated and uncultivated. In 
both cultivated and uncultivated plots nitrates under the maize were 
higher during July than in the bare soil. 



3 1 S JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Similar results have been obtained in our experiments in the field 
and also in lysimeters. In the lysimeter experiments records were 
kept of the quantity of nitrogen removed annually in the drainage 
waters and in the crops, the period for the calculations beginning 
May i and closing April 30 following. This covers the period during 
which the crop is on the soil and the interval before the next planting, 
during which time conditions are favorable for leaching out the ni- 
trates that may have accumulated in the soil during the summer. 
Comparing the quantities of nitrogen thus removed in the drainage 
water of implanted soil with the nitrogen in the crop plus that in the 
drainage water of planted soil we found that for maize the total quan- 
tity of nitrogen removed in the crop and drainage was considerably 
greater than from the unplanted soil. From oat soil slightly more 
nitrogen was removed than from bare soil, but from soil growing 
timothy and other true grasses less nitrogen was so removed. 

Taken as a whole these experiments indicate that with maize there 
is a stimulating influence on nitrate formation which is more potent 
than the depressing influence, while with timothy the opposite is the 
case. Oats appear to be intermediate between maize and timothy in 
their influence on nitrate formation and it is doubtful whether the 
stimulating or depressing effect is greater. 

Laboratory experiments in which methods were employed similar 
to those used to test the activities of nitrate-reducing bacteria were 
conducted with ammonifying bacteria. These gave some slight indi- 
cation that the solutions in which plants had grown produced some- 
what more ammonia from peptone than did similar solutions in which 
no plants had grown. Owing to the difficulty in getting pure cultures 
of nitrate-forming organisms the influence of plants on formation of 
nitrates has not been tried with these methods. 

Decomposition of organic matter is commonly and doubtless prop- 
erly regarded as one of the most important factors in rendering a soil 
fertile. If a crop e;m stimulate or retard this process at certain stages 
of its growth it holds the key that locks or unlocks the supply of plant 
nutrient for itfl own use and possibly influences to some extent the 
supply for the crops that follow. U has long been held that plants 
the power of rendering available for their own use the food 

ateriali contained in soils. The idea was advanced by Sachs that 

plant • 1 ■. 1 re t e organic acids which act on the inorganic matter 
fli olving ;i part of that which conies in immediate contact 
w ith the rOOl hair! and thus rendering it suitable for absorption. That 
organ' a< i'l other than carbonic are excreted by plant roots was 
5 '< >\ b\ the investigation^ of ( zapek and there arc few scien- 



LYON : INFLUENCE OF PLANTS ON SOIL BACTERIA. 3 I9 

tists who still believe that plants secure nutrients by such a process. 
The fact still remains that some plants apparently obtain more of 
certain inorganic substances from soils than would appear to be pos- 
sible from their solubility in water. Certain other plants obtain much 
smaller quantities of these nutrients. The influence which these plants 
exert on the decomposition processes in the soil may be a factor in 
determining whether they obtain much or little nutriment. 

Clover and alfalfa also appear to have a marked effect on bacterial 
activity in soils. It has, of course, been known from an early time 
that plants of this class increase the productivity of a soil both when 
plowed under and when raised for hay. The discovery some 30 years 
ago of the symbiotic relation of nitrogen-fixing bacteria and legu- 
minous plants apparently explained in full the reason for their useful- 
ness in promoting soil fertility. Since the growth of legumes on 
properly inoculated soil results in fixation of atmospheric nitrogen 
and therefore in an increase in the quantity of nitrogen in a soil the 
conclusion naturally follows that the resulting improvement in crop 
production is due to the increased supply of nitrogen that follows the 
growth of legumes. 

If, however, it is found that the growth of a legume has not in- 
creased the nitrogen content of a soil and yet that the productivity of 
that soil has been augmented, how then are we to account for the 
effect of the legume? An examination of the data at hand calls for 
the formulation of a different explanation. 

That the growth of a legume is not always accompanied by a greater 
accumulation of nitrogen in the soil than is the growth of certain 
native prairie grasses has been strikingly brought out by Swanson, 
who determined the nitrogen content of a large number of soils in the 
state of Kansas. Half of the samples were from alfalfa fields of 
from 20 to 30 years standing and the remainder had been in native 
grass pastures. In each case the alfalfa field and the pasture were 
near together and apparently of the same soil type. His results show 
that the alfalfa did not leave the soil any richer in nitrogen than did 
the other cropping treatment. Somewhat similar results were ob- 
tained by Alway and Bishop with soil from a clover field and a con- 
tiguous one in nonleguminous crops. 

At the Cornell experiment station four adjoining plots of land, two 
of which had been in timothy for six years and two in alfalfa for the 
same length of time, were carefully sampled and the nitrogen content 
of the soil determined. There was no difference between the timothy 
and alfalfa soil on one section and only 0.01 percent on the other. 
When these were plowed and planted to maize the alfalfa soil pro- 



320 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

duced about one-third larger crop of maize than the timothy soil. The 
legume had in this case increased the productivity of the soil as com- 
pared with timothy, but had accomplished this without leaving any 
larger nitrogen balance. 

Determinations of nitrates in the soil of these plots when kept free 
of vegetation and also in samples of the soil incubated after the addi- 
tion of dried blood showed a more rapid formation of nitrates in the 
alfalfa soil. There are two possible explanations for the greater 
nitrification in the alfalfa soil. Either the nitrogen left in the soil by 
the legume was more easily nitrifiable than the nitrogen in the timothy 
soil and likewise than that in dried blood or else the alfalfa plants had 
a stimulating effect on the activities of the nitrate-forming bacteria. 

In this experiment the roots of the alfalfa plants showed the pres- 
ence of tubercles and there was no reason to think that nitrogen fix- 
ation was not normal. Yields of the hay from these plots were esti- 
mated by weighing several, but not all, of the crops and these were 
analyzed for nitrogen. The figures show that there must have been a 
large fixation of atmospheric nitrogen by the alfalfa to supply this re- 
moval without depleting the store of soil nitrogen below that in the 
timothy soil. It seems probable that the more rapid formation of 
nitrates in the alfalfa soil, together with the well-known propensity 
of that plant to use nitrate nitrogen even when inoculated with nitro- 
gen-fixing bacteria, drew on the supply of accumulated nitrogen to 
such an extent that the final balance was no greater than it was in the 
timothy soil. 

I do not intend to question the well-known fact that the nitrogen 
content of a soil may sometimes be augmented by the growth of 
legumes even when the above-ground portions of the plants are re- 
moved, but I think it is fairly questionable whether the beneficial 
n- of legumes on soil productivity is due entirely to this increase 
in the nitrogen content or whether it is in part accomplished through 
some other influence which the plant exerts on the bacterial activities 
within the soil. 

These observations suggest the desirability of further investigating 
the conditions which determine the rate of accumulation and loss of 
-oil nitrogen attending the growth of legumes. It is well known that 

a librral supply of basic material in soil encourages the growth of red 
-love; ai • penally of alfalfa, but this condition of the soil also 
favors the formation of nitrate! which are absorbed in varying degree 
by the clover or alfalfa plants and arc likewise leached from the soil 
e percolating rainfall, h is que tionable, therefore, whether a 

well limed v oil would in the cud diow a larger balance of nitrogen fol- 



LYON : INFLUENCE OF PLANTS ON SOIL BACTERIA. 32 1 

lowing the growth of a legume than would a soil which has barely 
enough basicity to successfully produce the crop. 

Again the question arises as to the kind of legume best qualified to 
leave a large nitrogen balance in a soil. It is recognized that some 
legumes are capable of growing on more acid soils than are others. 
If the nitrogen-fixing power of all legumes were the same it is con- 
ceivable that the nitrogen balance resulting from their growth would 
be greater for those kinds that grow successfully on acid soil, because 
nitrate formation proceeds more slowly on such soils and the lesser 
use of nitrate nitrogen may compel a larger fixation of atmospheric 
nitrogen. From the point of view of nitrogen accumulation it may 
possibly be better for the farmer whose land is moderately sour to 
raise the legumes that are adapted to that soil rather than to lime it. 
In connection with this it now occurs to me that the Volusia silt loam 
of southern New York has a rather high content of organic nitrogen 
in spite of its slow rate of nitrate formation, while some of the more 
easily nitrifiable soil types contain less total nitrogen. However, we 
must admit that a large supply of nitrogen that is not available does 
not add to the value of a soil. It is possible that excessive basicity in 
a soil in a humid region is undesirable when considered from the 
standpoint of nitrogen economy. 

The experiments to which I have referred in this paper and the 
speculations in which I have indulged suggest another interesting and 
it may be useful line of investigation, namely, the effect of one crop 
on another growing in association with it. Mixed seedings of certain 
small grains have been found by several experimenters to yield more 
than either grown alone. Apple trees have often been found to grow 
poorly in sod land. The latter difficulty I think is in some cases 
clearly due to the depressing effect of the grass on the formation of 
nitrates in the soil and the consequent lack of nitrogen in the nutrition 
of the trees. 

Mixed seedings of a legume with a nonlegume often result in more 
vigorous growth of the nonlegume than when the latter is grown 
alone. Mixtures of the grasses with clover or alfalfa are likely to 
give especially good results, as the strongly depressing influence of the 
grass on nitrate formation is to some extent offset by the stimulating 
influence of the clover crop. There are doubtless a number of con- 
ditions that exercise an influence on the results of these mixed plant- 
ings and the nature and extent of these is a further subject of inves- 
tigation. It seems possible that plants like maize which themselves 
have a stimulating influence on nitrate formation do not benefit from 
association with legumes as do plants that depress nitrification. Lime 



322 JOURNAL OF THE AMERICAN' SOCIETY OF AGRONOMY. 

applied to soils on which lime is needed apparently increases the 
benefit to the nonlegume as does also the application of suitable fer- 
tilizer. 

The subject which I have outlined in a superficial way touches 
many aspects of crop production. Its study, however, is difficult. 
The complex nature of the soil and the rapid chemical and biological 
changes that occur when a soil sample is removed from the field or 
vessels in which plants are grown may entirely mask the effects pro- 
duced by plants. New methods must be devised and great patience 
must be exercised in the investigation of this subject. It is neverthe- 
less one of the problems that must be solved if we are to have the fun- 
damental knowledge on which to base a rational system of crop 
management. 

THE PREPARATION OF MANUSCRIPTS FOR PUBLICATION. 

C. W. Warburton. 

The preparation of manuscripts for publication in the Journal of 
the American Society of Agronomy or elsewhere deserves more 
than passing attention. If material is worth presenting at all, it is 
worth presenting in the best possible form. Carelessness or haste in 
the preparation of a paper often results in its rejection, even tho 
the matter it contains is otherwise worthy of publication. A poorly 
prepared paper, if published, is unsatisfactory to the writer and to the 
reade r, unless the editor devotes much time to its revision. 

Clearness and accuracy of expression are of major importance. 
X at u rally, not all men have the same ability to present facts in clear, 
concise language, but all can strive to obtain clarity of expression. 
Fine writing should be avoided. A short word is far better than a 
long one if it conveys the same idea quite as effectively. If six words 
can be made to do the work of ten, the omission of the useless verbiage 
i- a distinct gain. To quote George Otis Smith: 1 "The , . . scien- 
ti * ha^ at least two obligations: First, that of making his investiga- 
te, more and more exact in method and direct in result; second, 
that of making lii- product, the written report, such as to meet the 
noi only his professional associates but also the general 

public." 

It i not the intention, however, to make this brief note a discussion 
le, but rather one of form. Articles concerned with instruc- 

G >'C Otis. Plain writing. /;/ Science, n. s., 42: 630-632. 1915. 



WARBURTON : PREPARATION OF MANUSCRIPTS. 323 

tion, demonstration, experimentation, or research in agronomy will be 
accepted for publication in this journal. They may be reports of the 
results of original research, or they may be reviews of the work of 
others. Reviews of literature, however, should be critical digests of 
the available material on a subject rather than mere lists of titles. It 
is understood that articles submitted for publication have not been 
published previously elsewhere and that they will not be offered for 
simultaneous publication in other journals without the consent of the 
editor of the Journal of the American Society of Agronomy. 
Papers varying in length from I to 32 pages are acceptable; short 
papers will in general be given preference. 

Form of Manuscript. Articles for publication should be type- 
written on sheets approximately 7 by 1 1 inches in size. Carbon copies 
are not acceptable, as they are often blurred and are frequent sources 
of error, particularly in tabular matter. A duplicate copy should be 
retained by the author. Text pages should be double or triple spaced, 
preferably the latter, with wide margins. If the body of the text 
is triple spaced, double spacing may be used for quotations, citations, 
etc. Single spacing allows no opportunity for editorial changes and 
should never be used. The principal sections of the article should be 
indicated by subheadings, the relative rank of the subheadings being 
shown by underscoring or other means. Every page of a manuscript, 
including tables, should be numbered consecutively. 

Illustrations. Only such illustrations as are distinct additions to 
the text and aid in a clear understanding of it are acceptable. Each 
illustration must be specifically referred to in the text. Line drawings 
(text figures) are preferable to photographs, which require the making 
of half-tone engravings and printing on separate plates. Text figures 
should be numbered consecutively in the order of their occurrence, 
using Arabic numerals. Always refer to illustrations by number, as 
"figure 12," not "the following figure." If a distinct portion of a 
figure is referred to, it should be indicated by a capital letter in 'the 
text reference and also in the drawing itself, as " figure 14A." Text 
figures should be drawn in India ink on white or tracing paper, 
though the use of cross-section paper is permissible if it can not be 
avoided. All lettering should be clear and distinct. Each figure 
should be accompanied by a brief descriptive legend, plainly written. 

Photographs, when essential for use as illustrations, will be repro- 
duced as plates. Ordinarily two illustrations are reproduced on a 
single plate. Reference to plates should be by number, as " Plate 2, 
figure 1." Photographs for reproduction should be clear, black and 
white glossy prints. They should be unmounted, but should be 



324 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



attached slightly at the corners to a sheet of paper of the same size 
as the text of the manuscript. This sheet should bear the plate and 
figure numbers and a short descriptive legend. 

Tabular Matter. No other feature of editing occasions so much 
labor as the putting of tabular matter in form for publication. All 
table legends should be clear, concise, descriptive statements of the 
matter in the table. Tables should be numbered consecutively in 
Arabic numerals, and references to tables should always be by num- 
ber, as " Table 4," not " the following table." Arrange the data 
in the most compact form which will present clearly the information 
desired. Each column heading should indicate the nature of the data 
in that column. The unit of measure should be expressed in the 
legend if all data in the table are in the same unit, as "Dry weight 
in grams of crops grown, etc." If more than one unit is used, the 
unit for each column should be specified at the top of that column. 

Tables preferably should be written on separate sheets from the 
text matter. Care should be taken to limit the number of columns 
so that the table can be printed without difficulty on a page 4% 
inches wide. Otherwise, rearrangement and rewriting by the editor 
is usually necessary. If duplicate or replicate determinations were 
made, only the averages should in general be submitted for publica- 
tion, though the duplicate determinations may be sent for the editors' 
inspection. Footnotes to tabular matter should be designated by 
letters rather than index figures. 

Footnotes should be numbered consecutively throughout the paper, 
the first number being reserved for the date of receipt of the paper 
for publication and such other identifying statement as may be de- 
sirable. Copy for footnotes should be inserted in the text on the 
line immediately following the reference, and should be cut off by 
ruled lines above and below. If the text is triple spaced the footnotes 
may be double spaced. 

Citations of Literature may be printed as footnotes, though if they 
ral in number they may better appear at the end of the paper 
under the heading of "Literature cited." Citations so appearing 
-hould be arranger] in alphabetic order and should be numbered con- 

secutivi ' Reference in the text should be by this number, enclosed 

in parcntlicM-s. If more than one paper by an author is cited, the 
reference^ should be in the order of their publication, the earliest 
being dted first. Citation hould include the name of the author, 
'•vilh initial , litle of article, name of publication in which it appeared, 
with VOlun* and page if a periodical, and date of publication. Write 
n I% n before titles Of periodicals. Book citations should show place 



WAR BURTON '. PREPARATION OF MANUSCRIPTS. 



325 



of publication and, preferably, name of publisher. Note the follow- 
ing examples, and also recent issues of the Journal of the Amer- 
ican Society of Agronomy. 

11. Gardner, Frank D. Fertility of soils as affected by manures. U. S. Dept. 

Agr., Bur. Soils Bui. 48, p. 54. 1908. 
8. Briggs, L. J., and Shantz, H. L. The wilting coefficient for different 

plants and its direct determination. U. S. Dept. Agr., Bur. Plant Indus. 

Bui. 230, 83 p., 9 fig., 2 pi. 1912. 
1. Alvvay, F. J. Studies of soil moisture in the Great Plains region. In 

Jour. Agr. Sci., 4: 333-3^- 1908. 
3. Alway, F. J. Moisture studies of semiarid soils. In Rpt. 79th Meeting 

British Asso. Adv. Sci., p. 698, 699. 1908. 
14. Hopkins, C. G. Soil Fertility and Permanent Agriculture, p. 195. Ginn 

& Co., Boston. 1910. 

Figures and Abbreviations. Use Arabic numerals to express per- 
centages and measures of quantity or space, except at the beginning 
of a sentence, as "9 bushels," " 15 miles." Abbreviate metric weights 
and measures in all cases, but English weights and measures only 
when enclosed in parentheses, as "15 cm.," "45 bushels," but (45 
bu.). Use gm. for gram(s), cm. for centimeters (s), c. c. for cubic 
centimeter(s), kg. for kilogram(s). Use "percent," not "per cent," 
"per cent," or " %." 

Capitalization and Spelling. Follow Webster's New International 
Dictionary in capitalization and spelling. Spell " sulfur," " sulfate," 
and their compounds with " f ." Capitalize important words in cita- 
tions of book titles, but use small letters in citing titles of articles or 
bulletins. 

Proofs. Only one proof will be furnished. This should be read 
carefully, the necessary corrections indicated on the margin, and re- 
turned promptly to the editor. Do not make corrections in the text 
without indicating them also in the margin. Changes in figures, par- 
ticularly in tabular matter, should never be written over the originals. 
Make only absolutely necessary changes in the proofs. The original 
manuscript should be so written that radical changes in the proof are 
unnecessary. Proof, if unaccompanied by other matter, may be 
mailed at the rate of 1 cent for each 2 ounces. 

Reprints. Fifty reprints of each paper without covers are fur- 
nished free to authors. Additional reprints will be supplied at cost. 
Covers are supplied at the rate of $1.00 for the first 50, and 1 cent 
each for additional copies. Orders for reprints should always accom- 
pany the proof when it is returned. 



326 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



AGRONOMIC AFFAIRS. 

DELAY IN PUBLICATION OF THE JOURNAL. 

Because of the unavoidable postponement of the annual meeting 
of the Society, it has been necessary to delay publication of the De- 
cember number of this journal in order to include the annual reports 
of the officers and committees and the minutes of the annual meeting. 
The January, 1919, number has already been mailed and the succeed- 
ing numbers of volume 11 will be sent out on time, so far as possible. 

THE YEAR'S WORK. 

A.S stated in the annual report of the editor, printed elsewhere 
in this issue, the publication of the Journal of the American So- 
ciety of Agronomy during 1918 has not been accomplished without 
difficulty. The editor has been absent from his office during the 
greater part of the year, and as a consequence some of- the issues have 
been late in appearing. Not all have been up to the standard which it 
is desirable to maintain, while the exigencies of the times have made 
a reduction in the size of the annual volume necessary. The efforts 
of the members should now be devoted to the building up of the 
Society thru an increase in membership, thus making possible a larger 
and better publication in 1919. 

HONOR ROLL. 

As the entrance of several men into military service has been re- 
ported since the last previous issue of the Journal was published, it 
i- desirable to again print the Society's honor roll, so that the record 
may be as complete as possible. So far as known, 55 of the Society's 
members have been or arc now in military service, tho the list may 
-till be far from complete. Those whose names are known to the 
editor arc a^ follows : . . 



Aijikrt, A. K'., 
\m»kf.ws, Myron E., 
Buss, S. W., 

HWK KSON, W. I., 

Bruce, O. C, 
Brinson, A. M., 
Burnett, Groves, 

< ATI'S, Jlp.NRY R., 



( I! A I'M AN, JAMKS E., 

( Ihilm, EL R., 

I 1 UUEY, I 1 1 1< \ y E., 
I h KTXU k, K. P., 
Dl Wi RTF, 1 1. A., 
I )K KI.NSON, R, W., 
I )orr,i.AS, J. I'., 
Downs, K. K., 



Ellison, A. I)., 
Frkkman, Ray, 
Gentle, (i. E., 
Gilbert, M. B., 

( il<ATI AM, E. E., 

Gray, Samuel D., 

I Ialvkkson, W. V., 
Hanson, LewII P., 



Head, A. F., 
Helm, C. A., 
Holland, B. B., 
Hudelson, R. R., 
Jensen, O. F., 
Karlstad, C. H., 
Kenworthy, Chester, 
Kephart, L. W., 
Kime, P. H., 
Macfarlane, Wallace, 
Miner, Sterling, 



AGRONOMIC AFFAIRS. 

Moo MAW, Leroy, 
Newton, Robert, 
Palmer, H. Wayne, 
Piemeisel, R. L., 
Purington, James A., 

QUIGLEY, J. V., 

Ratliffe, Geo. T., 
Raymond, L. C, 
Richards, Phil E., 

SCHNEIDERHAN, F. J., 

NOTES AND NEWS. 



327 

schoonover, w. r., 
Scott, Herschel, 
Smith, J. B., 
Spencer, E. L., 
Stanley, C. W., 
Starr, S. H., 
Tabor, Paul, 
Towle, R. S., 
Ware, J. O., 
Westbrook, E. C. 



Roy O. Bridgeford has been elected instructor in agronomy at the 
Morris (Minn.) school of agriculture and F. W. McGinnis has been 
made instructor in farm crops in the same institution. 

L. A. Clinton, for the past several years assistant chief of coopera- 
tive extension work north and west in the U. S. Department of 
Agriculture, on November 1 succeeded Alva M. Agee as director 
of extension in New Jersey. Mr. Agee will continue as State com- 
missioner of agriculture. 

Howard S. Coe, assistant agronomist in the office of forage-crop 
investigations, U. S. Department of Agriculture, died at Beaumont, 
Texas, October 25, 191 8, of pneumonia following influenza. Mr. Coe 
was born at Orrville, Ohio, September 24, 1888. He graduated from 
Iowa State College in 19 13 and was granted the degree of M.Sc. by 
the same institution in 191 5. From September, 1913, to July, 1914, 
he was botanist and plant pathologist of the South Dakota station, 
resigning on the latter date to enter the service of the Department of 
Agriculture. His work during the past four years was for the most 
part in connection with studies of clover, sweet clover, velvet beans, 
and Southern pasture plants. He was the author of several bulletins 
of the Department of Agriculture and was a valued contributor to the 
pages of the Journal of the American Society of Agronomy, 
of which Society he was a member. Mr. Coe was married in Sep- 
tember, 1914, to Lela Marie Skinner, of Brookings, S. Dak., who, with 
a young son, survives him. He was an energetic worker of marked 
ability, and his loss is keenly felt by his associates. 

F. D. Farrell, for the past several years in charge of demonstra- 
tion work on reclamation projects for the Department of Agriculture, 
since September 1 has been dean and director of the Kansas college 
and station. 

L. H. Goddard has resigned from the States Relations Service, 
U. S. Department of Agriculture, to devote his entire time to his 



328 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



farming interests in Ohio. His work in connection with farm man- 
agement demonstrations has been taken over by the Office of Farm 
Management and will be directed by L. H. Moorhouse. 

Aven Nelson, for the past year acting president of the University 
of Wyoming, has been elected president. 

Ralph W. Redman, formerly of the States Relations Service, U. S. 
Department of Agriculture, is now assistant director of extension in 
Massachusetts. 

C. W. Stanley, after a year's service with the Canadian Expedi- 
tionary Forces, is now assistant analyst at the Ontario Agricultural 

College. 

F. H. Steinmetz has been elected assistant professor of farm crops 
and assistant agronomist of the Minnesota college and station and 
August Haedecke has been made assistant in agronomy of the same 

station. 

John H. Yoorhees has been added to the extension staff in farm 
crops at Cornell University. 

W. O. YVhitcomb, for the past five years assistant professor of 
agronomy at the Montana State College, has resigned to take charge 
of the Minneapolis office of the Seed Reporting Service, U. S. Depart- 
ment of Agriculture. 

Meeting of the New England Agronomists. 

The New England agronomists held their annual meeting in Boston, Novem- 
bcr 16, 1918. Those in attedance were G. E. Simmons of Maine; M. Gale East- 
man and F. W. Taylor of New Hampshire; A. B. Beaumont, H. P. Cooper, 
Earl Jones, and B. G. Southwick of Massachusetts; G. E. Adams of Rhode 
Island; and Henry Dorsey, B. C. Helmick, and W. L. Slate, jr., of Connecticut. 

An informal meeting was held on the evening of November 15 at the Parker 
House. Tin seed potato question was discussed, and Professor Slate outlined 
nil plan for inspection by the Food Administration of potatoes shipped into 
Connecticut. All those which do not meet the requirements for seed stock are 
sold for food During the discussion, it was suggested that the New England 
station- try certified seed potatoes in comparison with common northern grown 
seed. 

At die morning lesiion <>n November 16 the potato score card was discussed. 
Professor Adams told of his experiments with certified seed and of his work 
withp' t v ■ at the Khodr Island station. I 'rofessor Cooper of the Massachusetts 
then talked on the grading and judging of corn. The question of how 
to pick "ut the best samples in a corn show was discussed, and it was suggested 
' 11 features mighl be added to corn shows, such as economical produc- 
tion. Kerminatioii ti I . milling quality, etc. Mr. Southwick of the Massachu- 
OllegC talked; on the teaching of agronomy from the farm management 
standpoint. 

A commit!', appointed to draft the sentiments of the New England agrono- 



AGRONOMIC AFFAIRS. 



329 



mists on the value of corn shows and methods of conducting them, consisting 
of Jones, Southwick, and Dorsey, reported as follows: 

" In recognition of the fact that high-scoring corn as judged by our present 
scorecards may not necessarily be high-yielding corn, therefore, be it resolved 
that it is the sense of the New England agronomists that an attempt should be 
made to so modify our scorecards and other bases of comparison in judging 
corn that the factors of yield per acre and economical production may be given 
a larger recognition than at present, and that the chief object of corn shows 
should be to emphasize these two factors." 

On the passage of this resolution, the president appointed a committee to 
work on the problems brought out by the discussion, assigning a problem to 
each member, as follows: (1) To work out a plan for experimental work with 
good and poor corn as judged by the scorecard, Professor Adams; (2) to 
work pn the scorecard, Professor Cooper; and (3) to work on corn judging 
as related to boys' and girls' work, Mr. Eastman. 

Professor Beaumont of the Massachusetts station discussed soil surveys for 
New England. Following the passing of a resolution urging the taking of an 
inventory of the soil fertility resources of New England and one authorizing 
the appointment of a committee of three to bring the question of soil surveys 
before the meeting of experiment station directors of the New England, New 
York, and New Jersey stations, the chair appointed a committee for this pur- 
pdse consisting of Messrs. Beaumont, Simmons, and Slate. 

Following a discussion of teaching after the war, a resolution urging the 
adoption of a four-term system instead of the semester system by New England 
colleges was passed, with the suggestion that if the change is made the work 
in crop production should be scheduled during the growing season. The secre- 
tary was instructed to send copies of this resolution to the president's of each 
of the New England agricultural colleges. The secretary was also instructed 
to communicate with the New England Society for Rural Progress, with a view 
to the presentation of agronomic problems before that organization. 

Professor Slate was reelected president and Professor Jones secretary and 
treasurer for the ensuing year. 



330 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



REPORT OF THE SECRETARY-TREASURER. 



The present Secretarj'-Treasurer received his appointment from the Presi- 
dent, Dr. T. Lyttleton Lyon, under date of March n, 1918, following the resig- 
nation of Mr. P. V. Cardon. This report covers the period from that date to 
December 31, 1918, and includes the subscriptions, dues, etc., collected by the 
previous Secretary-Treasurer, subsequent to the filing of his report on March 
1. 1918. 

The year 1918 has been the most disastrous period in the history of the So- 
ciety. In previous reports there has always been recorded a substantial growth ; 
this one shows a decided decrease in membership. This decrease is due to 
several causes, perhaps most of all to the uncertainty of the positions of many 
of the younger members, owing to the draft. This has been an obstacle also 
in getting the usual number of new members. It is only fair to state, however, 
that the list of delinquents in the payment of dues contains the names of a 
number of the older agronomists. Whether these lapses were clue to careless- 
ness or were intentional, the Secretary can not say. There were some 260 mem- 
bers in arrears March 1. Since then two requests for payment have been sent, 
making three for the year. 

During the period covered by this report, 23 new members have been added to 
the Society, 2 have died, 5 have resigned, and 98 have been dropped for non- 
payment of dues. The Honor Roll of those entering the military service con- 
tains 55 names. It is quite likely that some who have been dropped for non- 
payment of dues belong on the Honor Roll. If so, the Secretary will gladly 
make due correction if his attention is called to the error. 

The Society has a paid-up membership of 509, including 10 members on the 
Honor Roll to whom the Journal has been sent. The names and addresses of 
-ix) members are printed elsewhere in this issue. Together with 45 others 
whose names arc included in the Honor Roll on page 326, the Society's member- 
ship at this time is 554. In addition to this membership list, there are 90 sub- 
scriptions to the Journal from libraries and other institutions. 

Financial Statement from March ii, 1918, to December 31, 1918. 



Receipts. 



\'< ■'•:•.< '! from |\ \ . Cardon, former Secretary 



$1,083.60 



I Jin- f p im members : 



\f>- members for i<;i8 

I member for 191 8 

I member for 1917-18 ., 

4 members for 191 9 

19 new members for 1918 
I new member for [9 iB 
I new member for m>i8 



at $2.50 $412.50 

at 2.00° 2.00 

at 2.00 4.00 

at 2.50 10.00 

at 2.50 47.50 

at 2.25'' 2.25 

at 2.00" 2.00 



Fifty cents still due to the Society. 



Agent's commission deducted. 



AGRONOMIC AFFAIRS. 



331 



5-oo 



2 new members for 1919 at 2.50 

2 student members for 1919 at 1.25 2.50 

13 local members for 1918 (N. C. section) at .50 6.50 

1 advance payment on 1919 dues 1.00 

Journal and Proceedings: 



II subscriptions for 1918 . 


at 


2.50 


27.50 


5 subscriptions for 1918 


at 


2.25 s 


11.25 


1 subscription for 1918 




2.40 6 


2.40 


12 subscriptions for 1919 


. . , , at 


2.25 s 


27.00 


1 subscription for 1919 


at 


2.50 


2.50 


Sale of volumes previous to 1918 






90.65 


Sale of reprints 






61.85 


Interest on bank deposit 






3.83 



Total recipts $1,805.83 



Disbursements. 

1918. 

March 25, Postage '. . $ 10.00 

April 1. Maurice Joyce Eng. Co 6.93 

April 13. Postage 10.00 

April 15. Maurice Joyce Eng. Co 1.50 

April 19. Mary R. Burr, clerical help 8.50 

May 18. New Era Printing Co 202.06 

June 10. New Era Printing Co 234.97 

June 14. Mary E. Burr, clerical help 5.00 

Jul}- 10. New Era Printing Co 311.96 

July 22. C. W. Warburton, postage, etc 12.00 

Aug. 1. Postage 8.00 

Aug. 1. Mary R. Burr, clerical help 5.00 

Aug. 30. New Era Printing Co 108.74 

Aug. 30. Maurice Joyce Eng. Co 48.69 

Sept. 4. New Era Printing Co ! i73-8o 

Sept. 5. Postage 10.00 

Oct. 3. Maurice Joyce Eng. Co 35-48 

Oct. 8. Postage ! 10.00 

Oct. 8. J. L. Wilson, postage, etc .65 

Oct. 29. Lewis M. Thayer, printing 5.75 

Nov. 23. Mary R. Burr, clerical help i7-8o 

Dec. 6. Maurice Joyce Eng. Co 4-5° 

Dec. 6. Lewis M. Thayer, printing 4-50 

Dec. 28. Postage 500 

Total disbursements $1,240.83 

Balance December 31, 1918 565-00 

$1,805.83 



Lyman Carrier, 
Secretary-Treasurer. 



332 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

By vote of the Society the report of the Secretary-Treasurer was adopted. 
The report of the Editor, as published elsewhere, was read and adopted. 
The report of the Committee appointed to canvass the votes on the amend- 
ment to Article 4 of the Constitution of the American Society of Agronomy to 
read. M The officers of The American Society of Agronomy shall be a President, 
a First Vice-President, a Second Vice-President, and a Secretary-Treasurer/' 
The committee, consisting of W. B. Ellett and A. B. Beaumont, reported 183 
votes for the amendment, 1 against, and 3 defective. The constitution was 
declared so amended. 

The nominating committee, consisting of C. A. Mooers and Robert Getty, 
reported the following nominations : 

President, J. G. Lipman, New Jersey Agr. Expt. Sta. 
First Vice-President, F. S. Harris, Utah Agr. Expt. Sta. 
Second Vice-President, A. B. Conner, Texas Agr. Expt. Sta. 
Secretary-Treasurer, Lyman Carrier, U. S. Dept. of Agriculture. 
By vote of the Society, these nominees were duly elected officers for the year 
1919. 

The report of the Committee on Standardization of Field Experiments, pre- 
pared by the Chairman, A. T. Wiancko, was read and adopted. 

On motion, it was voted that the Executive Committee consider the advisa- 
bility of affiliating with the American Association for the Advancement of 
Science and if thought desirable to leave the matter to a mail vote of the whole 
Society. 

Meeting adjourned. 

Address List of Members. 

Abell, M. F., College of Agriculture, Storrs, Conn. 
Adams, G. E., Agr. Expt. Sta., Kingston, R. I. 

Agee, John H., Bureau of Soils, U. S. Dept. Agr., Washington, D. C. 
Aicher. L. C, Aberdeen Experiment Farm, Aberdeen, Idaho. 
Albrecht, \Y. A., College of Agriculture, Columbia, Mo. 
Alexander, L L., State Normal School, Springfield, Mo. 
Allen. Edward R.. Agr. Expt. Sta., Wooster, Ohio. 
Allyn, Orr M., Fergus, Mont. 

Alvord, Emory D. ( Agr. Expt. Sta., Pullman, Wash. 

Alway, F. J., University Farm, St. Paul, Minn. 

Anderson, Arthur, University Farm, Lincoln, Nebr. 

Ap].. Frank, Rutgers College, New Brunswick, N. J. 

Amy. A. C, University Farm, St. Paul, Minn. 

Atkinson, Alfred, Agr. Expt. Sta., Bozeman, Mont. 

At water, ('. <;., Tin- Barrett Co., 17 Battery Place, New York, N. Y. 

0. L, Tenn. Coal and Iron Co., Birmingham, Ala. 
Babcock, F. R., County Agent, Crosby, N. Dak. 
Raelitell, M \. r >]•■«. Stale University, Columbus, Ohio. 
I'.ailey, ( . If., University Farm, St. Paul, Minn. 

Farm Management, U. S. Dept. Agr., Washington, 1). C. 
Ball, ( R . Bar riant Indus., U. S. Dept. Agr., Washington, D. C. 
!'..!»»< roft. Ross L., b.wa State < ollege, Ames, Iowa. 
Barbae, 0, K., Cliff House, Pullman, Wash. 
Itarrc, H. W., Ayi Expt Sta., ( Unison College, S. C. 



AGRONOMIC AFFAIRS. 



333 



Bartlett, Harley H., 335 Packard St., Ann Arbor, Mich. 
Bauer, F. C, 218 N. Lake St., Madison, Wis. 
Bear, F. E., College of Agriculture, Columbus, Ohio. 
Beaumont, A. B., College of Agriculture, Amherst, Mass. 
Beavers, J. C, Guilford College, N. C. 

Beeson, M. A., Oklahoma A. and M. College, Stillwater, Okla. 
Bell, Henry G., 11 11 Temple Bldg., Toronto, Canada. 

Bennett, Chas. D., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 
Bennett, Hugh H., Bureau of Soils, U. S. Dept. Agr., Washington, D. C. 
Berry, Roger E., 404 Knoblock St., Stillwater, Okla. 

Biggar, H. Howard, Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Billings, G. A., Farm Management, U. S. Dept. Agr., Washington, D. C. 

Birchard, J. F., Magnus Ave. and Main St., Winnipeg, Canada. 

Bizzell, James A., Cornell University, Ithaca, N. Y. 

Blackwell, C. P., Agr. Expt. Sta., Clemson College, S. C. 

Blair, R. E., Yuma Experiment Farm, Bard, Calif. 

Bledsoe, R. Page, Experiment Farm, Waterville, Wash. 

Block, J. F., Dept. of Interior, Calgary, Alberta, Canada. 

Bolley, H. L., Agr. Expt. Sta., Agricultural College, N. Dak. 

Boss, Andrew, University Farm, St. Paul, Minn. 

Boving, Paul, Univ. of British Columbia, Vancouver, B. C, Canada. 

Bower, H. J., Kansas State Agr. College, Manhattan, Kans. 

Boyack, Breeze, Agr. Expt. Sta., Fort Collins, Colo. 

Bracken, John, Saskatchewan Univ., Saskatoon, Sask., Canada. 

Brandon, Jos. F., Akron Field Station, Akron, Colo. 

Breithaupt, L. R., R. F. D. No. 3, Payette, Idaho. 

Brewer, Herbert C, The Barrett Co., 17 Battery Place, New York, N. Y. 
Briggs, Glen, Experiment Farm, Agana, Guam. 

Brodie, D. A., Farm Management, U. S. Dept. Agr., Washington, D. C. 

Brown, B. E., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Brown, E. B., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Brown, P. E., Iowa State College, Ames, Iowa. 

Bryant, Ray, Stillwater, Okla. 

Buckman, H. O.. Cornell University, Ithaca, N. Y. 

Bugby, M. O., Kingsville, Ohio. 

Bull, C. P., University Farm, St. Paul, Minn. 

Burdick, R. T., 635 N. California St., Stockton, Calif. 

Burgess, Jas. L., Dept. Agriculture, Raleigh, N. C. 

Burgess, P. S., Hawaiian Sugar Planters Expt. Sta., Honolulu, Hawaii. 

Burlison, W. L., Room 607 D. Agr. Bldg., Univ. of 111., Urbana, 111. 

Burnett, L. C, Iowa State College, Ames, Iowa. 

Burr, W. W., University Farm, Lincoln, Nebr. 

Burtis, Earl, 325 E. Olive St., Fort Collins, Colo. 

Bushey, A. L., Plankinton, S. Dak. 

Butler, Ormond R., Agr. Expt. Sta., Durham, N. H. 

Call, L. E., Agr. Expt. Sta., Manhattan, Kansas. 

Calvino, Mario de, Estacion Expt. Agron., Santiago de las Vegas, Cuba. 

Cardon, P. V., Judith Basin Expt. Farm, Moccasin, Mont. 

Carleton, M. A., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 



334 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Carr. Ralph H., 24 N. Salisbury St., La Fayette, Ind. 

Carrier. Lyman, Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Chambliss. Chas. E., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Champlin, Manley, Agr. Expt. Sta., Brookings, S. Dak. 

Chappelear. Geo. W., Normal Station, Harrisonburg, Va. 

Chilcott, E. F., Woodward Experiment Farm, Woodward, Okla. 

Childs. R. R.. College of Agriculture, Athens, Ga. 

Churchill. O. O.. Agricultural College, N. Dak. 

Clark, Chas. F., Box 747, Greeley, Colo. 

Clark. Chas. H., Bur. Plant Indus., Dept. Agr., Washington, D. C. 
Clark. J. Allen, Bur. Plant Indus., Dept. Agr., Washington, D. C. 
Clemmer, H. J., Woodward Experiment Farm, Woodward, Okla. 
Clevenger, C. B., University of Wisconsin, Madison, Wis. 
Cocke. R. P., Williamsburg, Va. 

Coffey, G. X., Extension Dept., Univ. of 111., Urbana, 111. 

John S.. Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 
Coleman. L. C, Director of Agr., Bangalore, Mysore, India. 
Conn. H. J.. Agr. Expt. Sta., Geneva, N. Y. 
Conner. A. B., Agr. Expt. Sta., College Station, Texas. 
Conner, S. D.. Agr. Expt. Sta., La Fayette, Ind. 
Conrey, ( i. \\\, College of Agriculture, Columbus, Ohio. 
Cooper, H. P., Mass. Agr. College, Amherst, Mass. 
Cooper, M. L., Merryville, La. 
Cowgill. H. B., Box 333, Fort Smith, Ark. 
Cowles, Henry C, University of Chicago, Chicago, 111. 
Cromer, C. Otis, Agr. Expt. Sta., La Fayette, Ind. 
Cron, A. B., Box 1214, Amarillo, Texas. 
Cunningham, C. C, Agr. Expt. Sta., Manhattan, Kans. 
Curtis, H. P., County Agent, Sutton, W. Va. 

Cutler, G. H., Univ. of Alberta, Edmonton South, Alt'a., Canada. 
Daane, \drian, 225 Duncan St., Stillwater, Okla. 
Damon, S. C, Agr. Expt. Sta., Kingston, R. I. 
Darst, W. H., Dept. Agr., State College, Penn. 

Davidson. Jehiel, Bur. Chemistry, U. S. Dept. Agr., Washington, D. C. 

R < > I Bureau of Soils, U. S. Dept. Agr., Washington, D. C. 
Dean, H. K.. Umatilla Experiment Farm, Hcrmiston, Ore. 
Deatrirk, Eugene I'., kntztown, Penn. 
Delwiche, E. J., 1221 Chicago St., Green Bay, Wis. 
Dillman, A. C, Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

iic, H. ('., A«r. Expt. Sta., Agricultural College, N. Dak. 
!' Henry, College of Agriculture, Storrs, Conn. 

Duley, F. L, 705 Hill St., Columbia, Mo. 
Dunnewald. T. J., 210 \. Carrol St., Madison, Wis. 
Dnnton, Leila, A«r. Expt. Sta., Manhattan, Kans. 
Dustman, Kobt. U„ Agr. Expt. Sta., Morgantown, W. Va. 
Dynes, 0. W . College of Agriculture, Ithaca, N. Y. 
EUett, W. B., A«r. Expt. Si.-... Blackfburgi Va. 

Oil, Paul, Agr. Expt Sta., College Park, Md. 
C C, College of Agric ill t lire-. New Hrunswick, N. J. 



AGRONOMIC AFFAIRS. 



Erdman, Lewis W., Agr. Expt. Sta., College Park, Md. 
Etheridge, W. C, Agr. Expt. Sta., Columbia, Mo. 
Evans, M. W., North Ridgeville, Ohio. 
Ewing, E. C, Scott, Miss. 

Fain, Jno. R., University of Georgia, Athens, Ga. 
Farrell, F. D., Agr. Expt. Sta., Manhattan, Kans. 
Fergus, E. N., Agr. Expt. Sta., La Fayette, Ind. 
Ferguson, A. M., Ferguson Seed Farms, Sherman, Texas. 
Fippin, E. O., Agr. Expt. Sta., Ithaca, N. Y. 
Firkins, Bruce J., Dept. of Soils, I. S. C, Ames, Iowa. 
Fisher, Forest A., Agr. Expt. Sta., Urbana, 111. 
Fisher, M. L., Purdue University, La Fayette, Ind. 
Fitz, L. A., Agr. Expt. Sta., Manhattan, Kans. 

Fletcher, C. C, Bureau of Soils, U. S. Dept. Agr., Washington, D. C. 

Florell, Victor FL, Plant Introduction Garden, Chico, Calif. 

Foesterling, H., Arbor Farms, Jamesburg, N. J. 

Foord, Jas. A., 54 Lincoln Ave., Amherst, Mass. 

Fraps, G. S., Agr. Expt. Sta., College Station, Texas. 

Frear, D. \V„ Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Fred, Edwin B., College of Agriculture, Madison, Wis. 

Free, E. E., 1105 Madison Ave., Baltimore, Md. 

Freeman, Geo. F., Societie Sultanienne d'Agriculture, Cairo, Egypt. 

Freeman, H. A., Experiment Farm, Ottawa, Canada. 

French, W. L., Austin, Minn. 

Furry, R. L., Ferguson Seed Co., Sherman, Texas. 

Gaddis, P. L., University Farm, Lincoln, Nebr. 

Gaines, E. F., Agr. Expt. Sta., Pullman, Wash. 

Galbraith, A. J., Manitoba Agr. Col., Winnepeg, Man., Canada. 

Garber, R. J., University Farm, St. Paul, Minn. 

Gardner, F. D., Agr. Expt. Sta., State College, Penn. 

Garren, G. M., P. O. Box 199, Raleigh, S. Dak. 

Garver, Samuel, Forage Crop Field Sta., Redfield, S. Dak. 

Gentle, G. E., Agr. Expt. Sta., Urbana, 111. 

Gericke, W. F., Agr. Expt. Sta., Berkeley, Calif. 

Gernert, W. B., Farm Bureau, Paris, 111. 

Getty, Robt. E., Branch Expt. Sta., Hays, Kans. 

Gilbert, M. B., Woodburn, Oregon. 

Gile, Philip L., 43 Briggs St., Melrose Highlands, Mass. 
Gillis, M. C, 401 E. Douglas St., Bloomington, 111. 
Gilmore, John W., College of Agriculture, Berkeley, Calif. 
Goddard, L. H., Washington Avenue, Washington C. H., Ohio. 
Goodall, S. E., Owensmouth, Calif. 

Gordon, Thomas B., Massachusetts Agr. College, Amherst, Mass. 
Graber, L. F., Wis. Agr. Expt. Assoc., Madison, Wis. 
Granovsky, Alexander, 320 Plum St., Fort Collins, Colo. 
Grantham, A. E., Agr. Expt. Sta., Newark, Del. 
Grantham, Geo, M., Agr. Expt. Sta., East Lansing, Mich. 
Grimes, W. E., Kansas State Agr. College, Manhattan, Kans. 
Gustafson, A. F., 962 State St., Ithaca, N. Y. 



336 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Hackelman. J. C. Agr. Expt. Sta., Columbia, Mo. 

Hagy. F. S., 924 Fremont St., Manhattan, Kans. 

Hall. Thomas D., College of Agr., Potchefstroom, South Africa. 

Hallsted. A. L., Branch Expt. Sta., Hays, Kans. 

Halverson, W. V., R. F. D. No. 2, Spanish Fork, Utah. 

Hansen. Dan, Huntley Experiment Farm, Huntley, Mont. 

Hardenburg, E. V., Dept. Farm Crops, Col. Agr., Ithaca, N. Y. 

Harlan, H. V., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Harlow. L. C. Agricultural College, Truro, N. S., Canada. 

Harris, F. S., Agr. Expt. Sta., Logan, Utah. 

Harrison. T. J., College of Agriculture, Winnipeg, Man., Canada. 

Hartley. C. P., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Hartwell, Burt L., Agr. Expt. Sta., Kingston, R. I. 

Haseltine, L. E., 2301 Durant Ave., Berkeley, Calif. 

Haskell. S. B., 1429 Munsey Bldg., Baltimore, Md. 

Hatcher, Otto, 318 West St., Stillwater, Okla. 

Hayes, Herbert K., University Farm, St. Paul, Minn. 

Headley, F. B., Truckee Carson Field Station, Fallon, Nev. 

Hearn. W. E., Bureau of Soils, U. S. Dept. Agr., Washington, D. C. 

Hechler, W. R., College of Agriculture, Ames, Iowa. 

Hendry, Geo. W., College of Agriculture, Berkeley, Calif. 

Hildebrand. E. B., 112 N. Main St., Stillwater, Okla. 

Hill, C. E., Experiment Farm, Moro, Ore. 

Hill, H. H., Agr. Expt. Sta., Blacksburg, Va. 

Hill. W. H., 326 Howe St., Vancouver, B. C, Canada. 

Hodgson, E. R., Agr. Expt. Sta., Blacksburg, Va. 

Holt, S. V., Experiment Farm, Rio, 111. 

Hopkins, E. S., School of Agriculture, Olds, Alberta, Canada. 
Hopt, Erwin, University Farm, Lincoln, Ncbr. 
Horton, Horace E„ 208 LaSalle St., Chicago, 111. " 
Hotchkiss, W. S., Substation No. 2, Troup, Texas. 
Hudelson, R. R., Agr. Expt. Sta., Columbia, Mo. 
Huelskemper, Ed., 411 Terry St., Longmont, Colo. 
Hulbert, H. W., University of Idaho, Moscow, Idaho. 
Humbert, Eugene P., Agr. Expt. Sta., College Station, Texas. 
Hume, A. X.. College of Agriculture, Brookings, S. Dak. 
lluiiKcrfon!, DcF., College of Agriculture, Fayettcville, Ark. 
Hunnicutt, B. II., Escola Agricola dc Lavras, E. de Minas, Brazil. 

I lurvt. J, B., 111 Knoblock St., Stillwater, Okla. 

Huston, H. A., German Kali Works, 43 Broadway, New York, N. Y. 
Hiitchcw,,!, T. ]',., A«r. Kxpt. Sta., I'.lacksburg, Va. 
I Iiitchisfifi. ('. B., College of Agriculture, Ithaca, N. Y. 

II • '.<<. S . Albert Dickinson Co., Chicago, 111. 
Mutton, ).(,, A«r. Kxpt. Sta.. I'.rookings, S. Dak. 

Juan U , VurimaKtias, Lonto, Peru, via Para, Brazil. 
Iko. Jerome, 228 W. Magnolia St., Fort Collins, Colo. 

on, Orson I-.. ( olk-Rc of Agriculture, Logan, Utah. 
FadctOtlj L D., Western ( auadl Floilf Mills Co., Winnipeg, Man. 
Jtcobf, Danirl ('., H. I. h \ M . f.Miyslmrg, Pa. 



AGRONOMIC AFFAIRS. 



Jardine, W. M., Kansas State Agr. College, Manhattan, Kans. 
Jarrell, J. F., Great Western Sugar Co., Longmont, Colo. 
Jenkins, J. Mitchell, Rice Experiment Station, Crowley, La. 
Jensen, L. N., Box 308, Big Springs. Texas. 
Johnson. D. R., 220 Knoblock St., Stillwater, Okla. 
Johnson, Geo. F., 256 \Y. Woodruff Ave., Columbus, Ohio. 
Johnson, T. C, Ya. Truck Expt. Sta., Norfolk, Va. 
Jones, Earl, Mass. Agricultural College, Amherst, Mass. 
Jones, J. W., Rice Field Station, Biggs, Calif. 
Jones, S. Agr. Expt. Sta., La Fayette, Ind. 
Joslyn, H. L., Farm Life School, Vanceboro, N. C. 
Justin, M. M., 442 State Capitol, Salt Lake 'City, Utah. 
Kan, F. F., P. O. Box 95. College Station, Texas. 
Karper, R. E., Lubbock Substation, Lubbock, Texas. 
Karraker, P. E., Agr. Expt. Sta., Lexington, Ky. 

Kearney, T. H., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 
Keim, Frank D., University Farm, Lincoln, Nebr. 

Kellerman, K. F., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 
Kelley, W. P., Citrus Experiment Station, Riverside, Calif. 
Kelly, E. O. G., Wellington, Kans. 
Kemp, W. B., Sparks, Md. 
Kennard, F. L., Colfax, Wash. 

Kennedy, P. B., 11 Budd Hall, Univ. of Calif., Berkeley, Calif. 

Kephart, L. W., Bur. Plant Indus., U. S. Dept. of Agr., Washington, D. C. 

Kezer, Alvin, Agr. Expt. Sta., Fort Collins, Colo. 

Kidder, A. F.. Agricultural College, Baton Rouge, La. 

Kiesselbach, T. A., Agr. Expt. Sta., Lincoln, Nebr. 

Kilgore, B. W., State Dept. Agr., Raleigh, N. C. 

Killough, D. T., Substation No. 5, Temple. Texas. 

Kinney, E. J., Agr. Expt. Sta., Lexington, Ky. 

Kirk. N. M., Bureau of Soils, U. S. Dept. Agr., Washington, D. C. 

Klein, Millord A., Spreckles Sugar Co., Spreckles, Calif. 

Klinck, L. S., University of British Columbia, Victoria, B. C, Canada. 

Knight. Chas. S., University of Nevada, Reno, Nev. 

Kraft, J. H., State Teachers College, Greeley, Colo. 

Krall, John A.. County Agent, Manchester, Iowa. 

Krauss, F. G., Supt. Extension Work, Haiku, Hawaii. 

Krusekopf, H. H., Agricultural College, Columbia, Mo. 

Kuska, J. B., Experiment Farm, Colby, Kansas. 

Langenbeck, Karl, 1625 Hobart St., N. W., Washington, D. C. 

Lapham, Macy H., Bureau of Soils, U. S. Dept. Agr., Washington, D. C. 

LaTourette, Lyman D., R. F. D. No. 1, Phoenix, Ariz. 

Laude, Hilmer H., R. F. D. No. 1, Beaumont, Texas. 

LeClerc, J. A.. Bureau of Chemistry, U. S. Dept. Agr., Washington, D. C. 

Ledyard, E. M., U. S. Smelting Co., 'Salt Lake City, Utah. 

Leidigh, A. H., Agr. Expt. Sta., College Station, Texas. 

Leighty, C. E., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Leith, B. D., University of Wisconsin, Madison, Wis. 

Leonard, Lewis T., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 



33* 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Lipman. C. B.. Agr. Expt. Sta., Berkeley, Calif. 
Lipman. J. G.. Agr. Expt. Sta., New Brunswick, N. J. 
Lippitt. W. D., 300 Sugar Bldg., Denver, Colo. 

Lohnis, F., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Loomis. Howard. S. Dak. State College, Brookings, S. Dak. 

Lora, Armando, Aguiar 47, Havana, Cuba. 

Love, H. H., Cornell University, Ithaca, N. Y. 

Love, Russell M., R. F. D. No. 2, Tarentum, Penn. 

Lowry, Marion W., College of Agriculture, Athens, Ga. 

Luaces, Roberto, Gran j a Escuela, Camaguey, Cuba. 

Lund, Viggo, Tystoftu Expt. Sta., Tjaereby, Denmark. 

Lynde, C. J., Macdonald College, Quebec, Canada. 

Lyon, T. Lyttleton, Cornell University, Ithaca, N. Y. 

McCall, A. G., Agr. Expt. Sta., College Park, Md. 

McCall. M. C, Experiment Farm, Lind, Wash. 

McClelland. C. K., Agr. Expt. Sta., Experiment, Ga. 

McDowell, C. H., Substation No. 6, Denton, Texas. 

McGuffey, C. Carl, McGuffey, Ohio. 

McHenry, Norris, R. F. D. No. 20, Elizabethtown, Ind. 

Mcllvaine, T. C, R. F. D. No. 4, Huntington, W. Va. 

McKee, Clyde, Iowa State College, Ames, Iowa. 

McKee, Roland, Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 
Mr Lane, J. W., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 
McMillan, S. A., Texas A. and M. College, College Station, Texas. 
M Miller, P. R., University Farm, St. Paul, Minn. 
Mi A'ess, G. T., Experiment Farm, Nacogdoches, Texas. 

Mackie, W. W., 121 Hilgard Hall, University of California, Berkeley, Calif. 
Madson, B. A., Agr. Expt. Sta., Davis, Calif. 
Maier, Fred, 400 S. Howes St., Fort Collins, Colo. 
Mally. F. W., Laredo, Texas. 

Maris, Edwin J., Demonstration Agent, Atwood, Kans. 
Martin, John H., Harney Branch Station, Burns, Ore. 
Maxson, A. C, Great Western Sugar Co., Longmont, Colo. 
Merklc, Fred G., Agricultural College, Amherst, Mass. 
M v - J. E., College Park, Md. 

Millar, C. E., Agricultural College, East Lansing, Mich. 

Miller, Edwin ('., Agricultural College, Manhattan, Kans. 

Miller. M. P., A^r. Expt. Sta., Columbia, Mo. 

Ifihon, Roy H., Clarksville, Tenn. 

Mii!<r. Sterling, 1029 Sixth St., Greeley, Colo. 

Miyake. Koji, College of Agriculture, Tohoku Imp. Univ., Sapporo, Japan. 

afonteaguda, Eieriberto de, Quinta de los ofoHnos, Habana, Cuba. 
Montgomery, K. G., College of Agriculture, Ithaca, N. Y. 
Mooers, Chas. A , Agr. Expt. Sta., Knoxville, Tenn. 
■ Harvey L, III N. Willow St., Trenton, N. J. 
H A., I'nivcrsity of Wisconsin, Madison, Wis. 

MoorhouM 1. \ , l«'ana ICanagament, U. s. Dept Agr., Washington, D, C. 

Morgan, J. 0., A«r Expt Sta., College Station, Texas. 
Morgan, O. S , Columbia University, New York, N. Y. 



AGRONOMIC AFFAIRS. 



339 



Morrison, J. D., Elbon, S. Dak. 

Morse, W. J., 6809 Fifth St., Takoma Park, D. C. 

Mortimer, Geo. B., College of Agriculture, Madison, Wis. 

Mosher, M. L., Eureka, 111. 

Mosier, J. G., Agr. Expt. Sta., Urbana, 111. 

Murray, Jas., Macdonald College, Quebec, Canada. 

Musback, Fred L., Marshfield, Wis. 

Myers, C. H., College of Agriculture, Ithaca, N. Y. 

Nash, C. W., Morris, Minn. 

New, T., Tsing Hua College, Peking, China. 

Newman, L. H., Canadian Building, Ottawa, Canada. 

Newton, Robert, Woodstock, New Brunswick, Canada. 

Oakley, R. A., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Olson, Geo. A., Agr. Expt. Sta., Pullman, Wash. 

Olson, P. ]., R. F. D. No. 1, Grafton, N. Dak. 

Osborn, L. W r ., Agr. Expt. Sta., Fayetteville, Ark. 

Osier, H. S., Court House, Ann Arbor, Mich. 

Pammel, L. H., Agr. Expt. Sta., Ames, Iowa. 

Park, J. B., College of Agriculture, Columbus, Ohio. 

Parker, John H., Agr. Expt. Sta., Manhattan, Kans. 

Pate, W. F., Agr. Expt. Sta., West Raleigh, N. C. 

Patterson, H. J., Agr. Expt. Sta., College Park, Md. 

Paxton, Glenn, P. O. Box 269, Fort Collins, Colo. 

Pendleton, Robert L., Ewing Christian College, Allahabad, India. 

Peterson, W T . A., Mandan, N. Dak. 

Petry, Edward J., 210 S. Ingalls St., Ann Arbor, Mich. 

Phillips, Thos. G., College of Agriculture, Columbus, Ohio. 

Pieters, A. J., 340 Blair Road, Takoma Park, D. C. 

Piper, C. V., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Pittman, D. W., Agr. Expt. Sta., Logan, Utah. 

Plummer, J. K., Agr. Expt. Sta., Raleigh, N. C. 

Pope, M. N., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Pridmore, J. C, 616 Rhodes Bldg., Atlanta, Ga. 

Rast, Loy E., State College of Agriculture, Athens, Ga. 

Reed, Everett P., Agr. Expt. Sta., Geneva, N. Y. 

Reed, H. R., Yuma Field Station, Bard, Calif. 

Reid, F. R., Bur. Plant Indus., U. S. Dept. Agr.. Washington, D. C. 
Reynolds, E. B., Substation No. 3, Angleton, Texas. 
Rice, Thos. D., Bureau of Soils, U. S. Dept. Agr.. Washington, D. C. 
Riley, J. A., Chester, S. C. 

Robbins, F. E., 432 Russell St., West La Fayette, Ind. 
Roberts, Geo., Agr. Expt. Sta., Lexington, Ky. 

Roberts, J. M., Central Scientific Co., 345 W. Mich. Ave., Chicago, 111. 
Ross, Jno. F., Cereal Field Station, Amarillo, Texas. 
Rost, C. O., University Farm, St. Paul, Minn. 

Rothgeb, B. E., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 
Roudebush, R. I., West Liberty, W. Va. 

Rueda, Buenaventura, Linea 7, Antiguo 8 y 10, Vedado, Havana, Cuba. 
Runk, C. R., 250 West 10th Ave., Columbus, Ohio. 



540 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Russell J. C, University Place, Lincoln, Neb. 

Ruzek, C. V., Agr. Expt. Sta., Corvallis, Ore. 

Salmon. S. C, Agr. Expt. Sta.. Manhattan, Kans. 

Salter, Robt. M., Agr. Expt. Sta., Morgantown, W. Va. 

Sarvis. T. T.. Northern Great Plains Field Sta., Mandan, N. Dak. 

Schafer, E. G.. State College, Pullman, Wash. 

Schaub. I. O.. States Rel. Service, U. S. Dept. Agr., Washington, D. C. 
Schmitz. Nickolas, Extension Dept., State College, Pa. 
Schoth. Harry A., Agr. Expt. Sta., Corvallis, Ore. 

Schreiner, Oswald, Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Schuer, H. W., 247 W. Tenth St., Columbus, Ohio. 

Schuster, G. L., Lancaster High School, Lancaster, Ohio. 

Scofield. C. S., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Seamans, Arthur, Huntley Experiment Farm, Huntley, Mont. 

Sears, O. H., 502 Waldron St., La Fayette, Ind. 

Sewell, M. C, College of Agriculture, Manhattan, Kans. 

Shantz. H. L., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Shaw, Chas. F., Agr. Expt. Sta., Berkeley, Calif. 

Shepperd, John H., Agr. Expt. Sta., Agricultural College, N. Dak. 

Sherwin. M. E., College of Agriculture, West Raleigh, N. C. 

Shutt, Frank T., Central Experiment Farms, Ottawa, Canada. 

Seiglinger, J. B., Woodward Field Station, Woodward, Okla. 

Simard. J. A., Box 211. Quebec, P. Q., Canada. 

Skinner, J. J., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Simmons, Geo. E., University of Maine, Orono, Maine. 

Slate. \V. L.. jr., Agr. Expt. Sta., Storrs, Conn. 

Slipher, John A., Agr. Expt. Sta., La Fayette, Ind. 

Smith. L. H., College of Agriculture, Urbana, 111. 

Smith, Nelson S., School of Agriculture, Olds, Alta., Canada. 

Smith. Ralph W., Dickinson Substation, Dickinson, N. Dak. 

Smith. Raymond S., 304 Elmwood Ave., Ithaca, N. Y. 

Smith. V. C. College of Agriculture, Columbus, Ohio. 

Snyder. Harry, [800 Summit Ave., Minneapolis, Minn. 

Southw ick. F j . G., College of Agriculture, Amherst, Mass. 

Southwick, Everett, 208 Lowell St., Peabody, Mass. 

Sotttbworth, W., Manitoba Agr. College, Winnipeg, Man., Canada. 

Spillman. \\ J.. Xorthbrook Courts, Washington, D. C. 

Spraggi F. A., Agr. Expt. Sta., East Lansing, Mich. 

Sqtrin . I ll.. Technical Div., Duponl Powder Co., Wilmington, Del. 

Stadler. L J., 410 S. Maple St., St. Louis, Mo. 

Stanley, Clarence W.. Ontario Agr. College, Cuclph, Out., Canada. 
Stanton, i . K , I'.nr. I'lant Indus., U. S. Dept. Agr., Washington, D. C 
Stemplc, I W., College of Agriculture, Morgantown, W. Va. 

Stephana, David E.. Bxperimenl Farm, Moro, Ore. 

Stephenson. R K., Iowa State College, Ames, Iowa, 
on. W. If., Agr. Expt. Sta., Ames, Iowa, 
rt, Georg< , \gr. Kxpt. Sta., Logan, Utah. 
Mr wart. II. W., 2010 Monroe St., Madison, Wis. 

Rnpeil L, < ollege of Agriculture, State College, N. Mcx. 



AGRONOMIC AFFAIRS. 



341 



Stoa, Theodore E., Agr. Expt. Sta., Agricultural College, N. Dak. 
Stockberger, W. W., Bur. Plant Indus., U. S. Dept. Agr., Washington D C 
Stokes, W. E., Edgefield, S. C. 
Stone, J. L., Agr. Expt. Sta., Ithaca, N. Y. 

Stookey, E. B., Western Washington Expt. Sta., Puyallup, Wash. 

Summerby, R., Macdonald College, Quebec, Canada. 

Taggart, J. G., School of Agriculture, Olds, Alta., Canada. 

Taliaferro, W. T. L., Agricultural College, College Park, Md. 

Taylor, F. W., Agr. Expt. Sta., Durham, N. H. 

Thatcher, L. E., Agr. Expt. Sta., Wooster, Ohio. 

Thatcher, R. W., University Farm, St. Paul, Minn. 

Thomas, Melvin, Charleston, 111. 

Thompson, G. E., Agr. Expt. Sta., Tucson, Ariz. 

Thompson, James, Col. of Pharmacy, Univ. of Wash., Seattle, Wash. 

Thompson, R. S., Highland, Calif. 

Thorne, Chas. E., Agr. Expt. Sta., W r ooster, Ohio. 

Throckmorton, R. I., College of Agriculture, Manhattan, Kans. 

Thysell, John C, Dickinson Substation, Dickinson, N. Dak. 

Tillman, B. W., University of Missouri, Columbia, Mo. 

Tinsley, J. D., 507 Union Station, Galveston, Texas. 

Tracy, S. M., Biloxi, Miss. 

Trout, C. E., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

True, R. H., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Truog, Emil, Agr. Expt. Sta., Madison, Wis. 

Turlington, J. E., University of Florida, Gainesville, Fla. 

Turner, A. F., College of Agriculture, Manhattan, Kans. 

Tuttle, H. Foley, College of Agriculture, Urbana, 111. 

Umberger, H. J. C, Extension Div., Col. of Agr., Manhattan, Kans. 

Van Alstine, Ernest, 912 Nevada St., Urbana, 111. 

Vanatter, P. O., College of Agriculture, Athens, Ga. 

Van Nuis, C. S., College Farm, New Brunswick, N. J. 

Van Schaick, K. L., Pretoria, Transvaal, South Africa. 

Veach, C. L., College of Agriculture, Athens, Ga. 

Veitch, F. P., Bur. Chemistry, U. S. Dept. Agr.. Washington, D. C. 

Viola, G. E., Oak and Spring Sts., West Hoboken, N. J. 

Vinall, H. N., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Vivian, Alfred, Ohio State University, Columbus, Ohio. 

Voorhees, John H., Extension Dept., Cornell Univ., Ithaca, N. Y. 

Waldron, L. R., Agr. Expt. Sta., Agricultural College, N. Dak, 

Waller, Adolph E., Ohio State University, Columbus, Ohio. 

Waller, Allen G., 2028 F St., N. W., Washington, D. C. 

Walster, H. L., Agr. Expt. Sta., Madison, Wis. 

Walter, E. J., 91 West Eleventh St., Columbus, Ohio. 

Walworth, E. H., Martinsville, 111. 

Warburton, C. W., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 
Welch, J. S., L. D. S. Maori Agr. College, Hastings, New Zealand. 
Wells, W. G., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 
Welton, F. A., Agr. Expt. Sta., Wooster, Ohio. 
Wentz, John B., Maryland State College, College Park, Md. 



54- JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Wermelskerchen, Louis, Agr. Expt. Sta., College Station, Texas. 

West, J. T., Agricultural College, Miss. 

Westgate, J. M., Agr. Expt. Sta., Honolulu, Hawaii. 

Westley, Roy, N. W. School of Agr., Crookston, Minn. 

Westover, H. L., Yuma Experiment Farm, Bard, Calif. 

Wheeler, Clark S., 423 W. Ninth St., Columbus, Ohio. 

Wheeler, H. C, Agricultural Bldg., Col. of Agr., Urbana, 111. 

Wheeler, H. J., 111 Grant Ave., Newton Center, Mass. 

Whitcomb, W. O., 320 Flour Exchange, Minneapolis, Minn. 

Whiting, A. L., 705 Gregory St., Urbana, 111. 

Whitson, A. R., Agr. Expt. Sta., Madison, Wis. 

Wiancko, A. T., Agr. Expt. Sta., La Fayette, Ind. 

Widtsoe, John A., University of Utah, Salt Lake City, Utah. 

Wiggans, Roy G., Cornell University, Ithaca, N. Y. 

Wilkins, F. S., College of Agriculture, Ames, Iowa. 

Wilkinson, T. V., 624 Egan St., Shreveport, La. 

Will, Geo. F., Bismarck, N. Dak. 

Willard, C. J., Ohio State University, Columbus, Ohio. 

Willey. Leroy D., Box 879, Sheridan, Wyo. 

Williams, Chas. B., Agr. Expt. Sta., West Raleigh, N. C. 

Williams, C. G., Agr. Expt. Sta., Wooster, Ohio. 

Wilson, Benj. D., Cornell University, Ithaca, N. Y. 

Wilson, Bruce S., Agr. Expt. Sta., Manhattan, Kans. 

Wimer, David C, Pa. State College, State College, Pa. 

Winters, X. E., Extension Dept., College of Agr., West Raleigh, N. C. 

Winters, R. Y., Agr. Expt. Sta., West Raleigh, N. C. 

Withycombe, Robert, Union, Ore. 

Wolfe, T. K., Agr. Expt. Sta., Blacksburg, Va. 

Wooward, John, Mt. Morris College, Mt. Morris, 111. 

Woods, A. F., Maryland State College, College Park, Md. 

Woodworth, C. M., Agr. Hall, Univ. of Wis., Madison, Wis. 

Worsham, W. A., College of Agriculture, Athens, Ga. 

Wright, A. H., Dept. of Agronomy, College of Agr., Madison, Wis. 

Wright, R, Claude, Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Wnnsch, W. A., County Agent, Newton, Kans. 

Wyatt. P. A., 216 Agr. Bldg., University of Illinois, Urbana, 111. 

Voder, P. A., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

YounK, Philip U., Agr. Expt. Sta., Manhattan, Kans. 

Yotingblood, Bonney, Agr. Expt. Sta., College Station, Texas. 

/.avitz, C. A., Ontario Agr. College, Guelph, Ont., Canada. 

Zinn, Jacob, Agr. Expt. Sta., Orono, Maine. 

/00k, L. L., North Platte Substation, North Platte, Nebr. 

La »' i.i) M km itr.Rs. 

The following have lapsed for non-payment of 1918 dues: 
Aim ms, E. L., BOAtDMAJT, W. C, P.ootii, V. J., 

I'.AkKU lo ) I'll I , BOILAND, J., BoilYOUCOS, S. J., 

' i" 1., Bovazzx, A., Caxsoll, J. S., 

1 i Bom m it, R, EL < artkr. l. M.. 



AGRONOMIC AFFAIRS. 



343 



Collins, M. W. H., 
Cook, I. S., Jr., 
Cox, Jos. R, 
Crabb, Geo. A., 
Cramer, W. F.. 
Criswell, Judson, 
Crosby, M. A., 
Davison, W., 
Derr, H. B., 
Dickson, R. E., 
Doryland. C. J. T., 
Dougall, Robert. 

DU BUISSON, J. P., 
DUGGAR, J. F., 

Ellis, Orland I., 
E wan. A. E.. 

FlNNELL. H. H., 

Fleming. Frank L.. 
Fletcher, O. S., 
Foreman, L. W., 
Frank. W. L., 
Gilkerson, H. C, 
Gish, N. A., 
Guell, Aurelio R., 
Hanger, W. E., 
Hanson, H. P., 
Harrington, O. E., 
Haskell, E. S., 
Hill, P. R., 



Hodson, Edgar A., 
Hoke, Roy, 
Holbert, J. R., 
Hopkins, Cyril G., 
Hurst, J. B., 
Hutchinson, W. L., 
Hyslop, Geo. R., 
Jackson, J. W., 
Jarvis, Orin W., 
Kemp, Arnold R., 
Kenney, Ralph, 
Knutson, George, 
Lechner. H. J., 
Letteer, C. R., 
Livingston, Geo., 
Long, David D., 
Longman, O. S., 
Luckett, J. D., 
McClymonds, A. E., 
McIntire, W. H., 
Marbut, C. F.. 
Martin, Thos. L., 
Mathews, Oscar R., 
Maughan, Howard ]., 
Miller, Frank R., 
Milner, F. W. s 
Morison, A. T., 

MOYNAN, J. C, 
MUNDELL, J. E.. 



Murphy, Henry, 
Nevin, L. B., 
Newman, C. L., 
Noyes, H. A., 
Olson, M. E., 
Osenbrug, Albert, 
Prince, Ford S., 
Pritchard, F. J.. 
Richey, F. D., 
Robertson, R. B., 
Rose, C. N., 
Schultz, H. H., 
Severance, George, 
Shinn, E. H., 
Shoesmith, V. M., 
Sink, Stanley B., 
Smith, Howard C, 
Snyder, Robert M., 
Stewart, Robert, 
Torgerson, E. F., 

TOWNSEND, C. O., 

Van Ever a, R., 
Wallace, R. C. E., 
Ward, Wylie, 
Wascher, F. M. W.. 
Winsor, L. M., 
Win wright, George, 
Woo, Moi Lee. 



544 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



MINUTES OF THE ELEVENTH ANNUAL MEETING. 

Baltimore, Md., January 6-7, 1919. 

First Session, Monday Afternoon, January 6. 

The meeting was called to order by the Secretary-Treasurer in the absence of 
President Lyon and Prof. J. H. Shepperd was appointed Chairman. The fol- 
lowing papers were presented : 

1. Effect of Varying Degrees of Heat on the Viability of Seeds, by James L. 
Burgess (followed by a discussion). 

2. Field Crop Inspection, a Necessity to Standardization and Crop Improve- 
ment in Cereals, by H. L. Bolley (read by J. H. Shepperd). 

Second Session, Monday Evening, January 6. 
Dr. Herbert Osborn, presiding. 

Some Observations on Agricultural Conditions in England and France, by 
Dr. W. O. Thompson. 

Influence of Higher Plants on Bacterial Activities in Soils, by Dr. T. L. Lyon, 
President of the American Society of Agronomy (read by Dr. C. E. Leighty). 

The Problems of Permanent Pasture with Special Reference to its Biological 
Factors, by Dr. Herbert Osborn, President of the Society for the Promotion of 
Agricultural Science. 

Third Session, Tuesday Morning, January 7. 

3. The Small Grain Varieties of Utah, by George Stewart (read by F. S. 
I [arris). 

4. I-Vrtilizi r Kxperiments on DeKalb Soils, by Frank D. Gardner. 

1 rreen 6and Deposits as a Source of Potassium for Crops, by R. H. True. 

6. Carrying Capacity of Native Range Grasses, by J. H. Shepperd. 

/ ourth Session, Tuesday Afternoon, January 7. 

7. \ Method for Determining the Proper Stand of Corn under Southern 
( r.mlitions. by C. A. Mooers. 

X. Thi Work of the Committee on Seed Stocks, by R. A. Oakley. 

g \ 1 foi the Contradictory Results in Corn Kxperiments, by Lyman 

Carrier. 

Buthiist Meeting, 

1 report of the Secretary-Treasurer, ;is presented elsewhere in this issue, 

wa* read. 

Th| report of the Auditing Committee was then read by the Chairman, Frank 

D. Gardner, a» follows: 



AGRONOMIC AFFAIRS. 



345 



Report of Auditing Committee. 

Your committee has audited the accounts of the Society by the Secretary- 
Treasurer, Lyman Carrier, and find them to be correct. 

(Signed) Frank D. Gardner, 
C. F. Marbut, 

Committee. 

REPORT OF THE COMMITTEE ON STANDARDIZATION OF 
FIELD EXPERIMENTS. 

In the work of the Committee on the Standardization of Field Experiments, 
it was thought advisable this year to make a survey of the methods actually 
followed by investigators employing field plot experiments. The task of col- 
lecting this information was divided among the members of the committee, one 
taking the questions relating to size, shape, and arrangement of plots em- 
ployed in soil-fertility investigations ; another taking similar questions relating 
to field experiments with crops, and the third taking the questions regarding 
the use and management of check plots. In each case questionnaires were pre- 
pared and sent to the workers along these lines in the experiment stations 
thruout the United States. It was felt that the information thus gathered 
would be of value to the members of the Society and would be helpful later 
on in formulating some general rules for the guidance of experimenters plan- 
ning new work along these lines. 

FIELD-PLOT METHODS IN SOIL-FERTILITY INVESTIGATIONS. 

Size of Plot. — The answers to the question regarding the size of plots em- 
ployed in soil-fertility experiments showed that nearly all of the experiment 
stations are using several different sizes of plots. There are various reasons 
for this. In many cases the area of land available has been the determining 
factor. Sometimes the nature and extent of the particular experiment has 
influenced the size of plot employed. More often the ideas of the experimenter 
as regards the most suitable and convenient size of plot have been the deter- 
mining factor and so successive workers making additions to field-plot experi- 
ments have laid out plots of different sizes. Even tho the work is quite similar 
in nature, we thus have sometimes as many as a half dozen different sized plots 
in the same experiment field. 

Among the 29 stations answering the question, the sizes of plots employed 
vary from 1 acre down to 1/200 of an acre. It was not possible to tell from the 
answers just how many different sizes of plots are employed, because in many 
cases the question was answered by merely naming the extremes of variation 
in size. Neither was it possible to get an accurate idea of the average size of 
plots used, but it is probably somewhere between one-tenth and one-twentieth 
acre. Only one of the 29 stations is- using plots as large as 1 acre or as small 
as 1/200 acre. Tenth-acre and twentieth-acres plots largely predominate, and 
fortieth-acre plots appear to stand next in numbers. Only a few stations are 
using plots larger than tenth-acre or smaller than fortieth acre. A few of 
the stations are using quarter-acre, fifth-acre, eightieth-acre, or hundredth-acre 
plots. In answer to the question, " What do you consider the ideal size of 
plot?", the great majority suggested the twentieth-acre and tenth-acre sizes. 
Only a few favored either larger or smaller plot's for ordinary purposes. 



346 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Shape of Plot— Twenty-two of the twenty-nine stations more or less defi- 
nitely answered this question and it appears that the variations in shape are 
even greater than the variations in size. Nearly every experimenter has a 
different idea as to the best or most convenient shape of plot. In 20 of the 22 
cases where shape of plot is specified, more or less rectangular plots are used, 
and in the majority of cases they are decidedly long and narrow. Variations 
run from twice to ten times as long as wide. In two cases square plots are 
used. 

It was brought out in many of the replies that both size and shape of plot 
must necessarily vary according to the acreage and character of the land avail- 
able and that no fixed system can be followed in all cases. 

Arrangement of Plots to Overcome Variations in Soil Fertility. — The im- 
portance of arranging series of plots so as to equalize soil differences as much 
as possible is generally recognized by experienced investigators. A common 
practice is to lay off long, narrow plots at right angles to the main soil varia- 
tions and to provide for making corrections for the variations in the other 
direction by the use of frequent uniformly treated check plots. The repetition 
of series of plots so as to have a similarly treated series for each crop in the 
rotation helps to equalize variations. Such repetition of series provides for 
overcoming the effects of seasonal variations upon different crops by making 
it possible to grow the several crops employed in the rotation every year. This, 
together with frequent checks and averaging the similarly treated plots in the 
several series over a period of years, is generally considered sufficient to over- 
come ordinary variations. 

Replication of Trials. — On this point, it was not possible from the answers 
to tell to what extent replication is practiced. The question was evidently in- 
terpreted in two different ways, some taking it to mean the number of similarly 
treated sets of plots and others the number of times the whole experiment is 
repeated. In most cases, however, each series of plots is repeated as many 
times as there are crops in the rotation followed, as for example, a 4-year 
rotation of corn, oats, wheat, and hay would have four sets of similarly treated 
plots, one set for each crop and all crops grown every year. Actual replication 
of whole series of experiments is practiced by only a few of the stations, tho 
several expressed themselves in favor of it wherever practicable. However, 
the large amount of land and labor required is a practical difficulty involved 
in the frequent replication of field-plot experiments, which usually include a 
considerable number of plots. 

1} Borders. As regards spaces between plots and borders around 
series, the practice is to have narrow untreated strips between plots and un- 

treated borderi around leriea, planted to the same crop and cut out or trimmed 
off at harvest time. Only a few considered spaces and borders unnecessary. 

/ '<•'/"' ■ nf ( link finis. In answer to the question, "What is your prac- 
tice as regards the frequency of check plots?", 33 answers were received. It 
was found that many of the stations did not follow any one standard system 
of arranKitiK check plotl and mat in many cases the practice varies according 
to the particular exp< 1 inn nt. At the same station, series of plots laid out at 
may htVC checks located at different intervals. Thirteen of the 
Minn thr question tlni have more than one system in operation. 

iblr from iln answer lo tell how many different systems 



AGRONOMIC AFFAIRS. 347 

are in use. In the following classification of the number of stations using 
different frequencies of checks, some overlapping occurs, as some are using 
more than one system. At 4 of the stations, systems of plotting are used in 
which every other . plot is a check. The system of having every third plot a 
check is used at 21 of the stations. Six of the stations reporting are using 
every fourth plot as a check; 10, every fifth; 3, every sixth, and 4, every tenth. 
One of the stations is using a check near the beginning and near the end of 
each series of plots. In one case only one check is used, located at the begin- 
ning of each series, and at two of the stations the checks are located at irregu- 
lar intervals according to convenience or soil variation. In many cases, perhaps 
the majority, the field-plot experiments reported were laid out years ago and 
a considerable number of them were laid out shortly after the establishment 
of experiment stations in this country, when the men in charge had little expe- 
rience back of them. In making suggestions for laying out new series of per- 
manent plots for soil-fertility investigations, the great majority of those reply- 
ing favored the use of frequent and regularly distributed checks. In most 
cases it is considered desirable to have a check plot on one side or the other 
of each specially treated plot. 

Treatment of Check Plots. — To the question, " How should check plots be 
treated?", 32 replies were received. Here again the practice varies, some 
giving the check plots a standard maintenance treatment, but in the majority 
of cases the check plots receive no treatment in the way of additions to the 
soil other than the crop rotation. A few of the stations are using treated 
check plots in some experiments and untreated check plots in others. In classi- 
fying the replies to this question, it was found that the answers were not always 
definite. At least 7 of the stations are giving the check plots some definite 
and regular treatment, while 18 are using totally untreated check plots only, 
altho 9 of them think it might be advisable to give the checks some regular 
treatment calculated to keep them in a normal state of productiveness. In 
discussing the question from the standpoint of future work, 24 stations out of 
32 answering the questions favor giving the plots which are used merely for 
checking purposes a uniform standard manurial treatment calculated to keep 
them in a reasonably productive condition, provided that one or more untreated 
plot's are included in each series to show what will happen to the land if crops 
are continually removed and nothing returned. The most commonly suggested 
treatment is a uniform dressing of manure. Some favor liming only, others 
a green manuring at certain regular intervals. Many of the workers along 
this line feel that the totally untreated plot becomes more and more unsatis- 
factory for checking purposes as the fertility runs lower and lower, and the 
results of the special treatments studied become unduly exaggerated. In a 
few cases, it is held that it is not possible to devise a system of treating check 
plots that will maintain uniformity, and it is believed that difficulties might 
arise which would be more serious than those connected with totally untreated 
checks. 

Methods of Computing Results Through Check Plots. — To the question, 
" How should check plots be used in computing the results of field experi- 
ments?", 34 replies were received, but 6 of them were not definite enough to 
admit of classification. Twelve investigators of the twenty-eight definitely an- 
swering the question use the check plots by calculating the normal check yield 



343 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



for each of the plots lying between the actual checks by assuming that the soil 
changes gradually from one check plot to the next. The difference between 
this calculated check yield and the actual yield is taken as the effect of the 
special treatment given to the plot. In 5 cases, the practice is to compare the 
specially treated plots with the average of the two nearest checks. In 3 cases, 
each specially treated plot is compared with the average of all the checks in 
the series. In 1 case where every third plot is a check, the treated plot is 
compared directly with its adjoining check. In 3 cases, sometimes the first 
system and sometimes the second system is used. In 1 case, sometimes the 
first system and sometimes the third is used, and in 2 cases, sometimes the 
second system is used and sometimes the third. In 1 case, either the third 
or fourth system is used and in another, any one of the four systems is used, 
according to conditions. In 1 case, the probable error is computed for each 
plot and stated after the actual yield for that plot. From the foregoing it is 
seen that here again there is no uniformity among the stations in the system 
employed, some using one and some another of four or five different systems. 
Check plots are used in one way or another by all of the stations answering 
the questions. However, it is doubtful if there is justification for so many 
different practices. Alt'ho actual practices differ so widely, the preponderance 
of opinion is in favor of the use of frequent and regularly distributed checks, 
giving the check plot some regular soil treatment to keep it in a condition of 
reasonable productiveness and employing the check in making corrections by 
calculating the probable check yield for each plot according to the variation 
between the two nearest checks. 

METHODS OF CONDUCTING EXPERIMENTS WITH CROPS. 

A questionnaire relating to the methods used in conducting experiments with 
crops was sent to 76 agronomists. The principal questions related to the 
methods generally used in varietal, rate-of-seeding, date-of-seeding, method-of- 
seeding, and stage- or time-of-harvesting tests with small grain, cultivated 
crops, and hay and pasture crops in both field and nursery. Thirty-six replies 
to a part or all of the questions were received. These replies are summarized 
for each of the larger groups of crops as follows: 

SiSi and Dimensions of Plots Used for Varietal, Rate-of -Seeding , and 
Date-of -Seeding Tests with Small Grain. — Thirty investigators reported the 
plots used for varietal, rate-of-seeding, and date-of-seeding tests with 
small grains. Of these, 2 use (not exclusively) plots larger than one-tenth acre 
/• Twenty use plots varying in size from one-twentieth to one-tenth acre, 
12 USC plots less than one-twentieth but larger than one-eightieth acre, and 2 
DSC plots One-eightieth acre in Size or smaller. One investigator who uses 
fortieth-acre plots states that unpublished work indicates that one-eightieth 
acre is a good size. 

reports relating to varietal, rate-of-seeding, and datc-of-sccding 
testl v.ith rowed crops (corn, sorghum, cotton, cowpcas, soybeans, and pota- 
received. Of these, 3 use plots larger than one-tenth acre, 18 use 
' i 01 H '.nth to one-twentieth acre, [6 use plots less than one-twentieth 

but larger than one-eightieth acre, and 5 use plots less than one-eightieth acre 
in size. 

I v. < nty-ninc investigators reported on varietal and rate- and date-of-seeding 



AGRONOMIC AFFAIRS. 



349 



tests with hay and pasture crops (alfalfa, clover, and grasses). Of these, 7 
use plots larger than one-tenth acre, 21 use plots from one-twentieth to one- 
tenth acre, 8 use plots less than one-twentieth and more than one-eightieth acre, 
and 4 use plots one-eightieth acre in size or smaller. Most of those who use 
plots larger than one-tenth acre do so for pasture experiments. 

Twelve reported miscellaneous tests such as stage of cutting and time of 
harvesting in which different sizes than those reported for other tests were 
used. Of these. 3 use plots larger than one-tenth acre, 8 use plots one-twentieth 
to one-tenth acre, 3 use plots less than one-twentieth acre but larger than one- 
eightieth acre, and 2 use plots one eightieth acre in size or smaller. 

Summarizing all crops and kinds of experimental tests, it appears that slightly 
more than half of the agronomists use plots varying in size from one-twentieth 
to one-tenth acre. Very few use plots larger than one-tenth or as small as 
one-eightieth acre. For small grains and rowed crops, plots less than one- 
twentieth but larger than one-eightieth acre are used almost as much as the 
one-twentieth to one-tenth acre sizes. Plots larger than one-tenth acre are 
used more extensively for hay and pasture crops than for small grains or 
rowed crops. 

Thirty-one reported the number of rows per plot for tests with rowed crops. 
Of these 8 use plots consisting of one row, 12 use plots consisting of two or 
three rows, and 27 use plots consisting of four or more rows per plot. 

All experimenters who gave dimensions reported the use of long, narrow 
plots, the ratio of width to length varying from about 1 to 4 to about 1 to 200. 
The average ratio for all reporting is about 1 to 28 for small grains, 1 to 15 for 
rowed crops, and about 1 to 10 for hay and pasture crops. 

Number of Times Plots are Replicated. — Thirty-three replies relating to va- 
rietal tests with small grains were received. Six of these report the use of 
single plots. Twenty-one duplicate some or all of their tests, 17 replicate from 
two 1 to four times, and 5 replicate their tests more than four times in all cases 
or when time and ground permit. For rate- and date-of-seeding tests, 3 ex- 
perimenters use single plots, 17 duplicate their tests, 14 replicate two to four 
times, and 3 replicate more than four times. 

For varietal tests with rowed crops, 7 investigators report the use of single 
plots, 20 duplicate their tests, 16 replicate two to four times, and 2 more than 
four times. For rate- and date-of-seeding tests with rowed crops, 5 use single 
plots, 13 duplicate, 9 replicate two to four times, and 1 more than four times. 

Of those reporting tests with pasture and hay crops, 8 use single plots. 15 
duplicate their tests, 9 replicate from two to four times, and 1 replicates more 
than four times. A number who used single plots indicated that they did so 
only when land and facilities for handling the work made it impractical to do 
otherwise. In fact, only one investigator reported the use of single plots ex- 
clusively. In this case and in most of the others reporting the use of single 
plots, check plots were used. 

In interpreting results of experiments in which replicate plantings are made, 
16 compute the " probable error " and 13 do not. 

Margins of Plots. — An inquiry was made as to whether the margins were 
harvested with the plots, removed before harvest, or whether the alleys and 
roadways were seeded solid to prevent the crop receiving water and plant food 

1 That is, three plots of each variety. 



350 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



from the adjoining areas. Thirty-four replies were received. Thirteen re- 
ported that the outside rows or margins were harvested and included with the 
plots. Fifteen do not include the outside rows or margins with the plot; i.e., 
the outer .edges are removed before harvest. Six follow the practice of seed- 
ing the entire alleys and removing the crop before harvesting the plots. 

Some or all of those who make no provision for eliminating the marginal 
effect use very narrow alleys and one includes a part of the alley with the 
plots in computing yields in order to compensate for the food and water se- 
cured from beyond the edges of the plots. One investigator in tests with small 
grains seeds winter wheat in the alleys in the spring. 

The Colorado station has determined for irrigated plots the lateral move- 
ment of water and surrounds the plots with borders which extend beyond 
the limit of safety. 

Twenty-five replied as to provisions being made to prevent insect or fungus 
damage to adjoining .plots because of the proximity of especially susceptible 
varieties, dates of cutting (as, for example, leaf spot on alfalfa, which may 
spread from a plot which is not cut to the new succulent growth of an adjoin- 
ing plot and do more damage than if the entire field was cut at the same time) 
or special methods of seeding or cultivation. Fifteen reported that no such 
precautions were taken. The remainder endeavored to prevent such damage 
by grouping susceptible varieties, by eliminating susceptible varieties as soon 
as discovered, by early harvest of severely damaged plots, and by control 
measures. 

Thirty-two agronomists reported regarding provisions made to prevent errors 
ill experimental tests with rowed crops due to the competitive effect of adjoin- 
ing plots, as, for example, shading by unusually tall varieties of corn, or early 
planted lots in time-of-seeding tests. Fifteen reported that no provisions of 
this kind arc made. Five of the remainder group the varieties, and ten elimi- 
nate or reduce errors of this kind by planting extra or guard rows. 

Variation in Sice of Kernels in Varietal Tests. — The question was asked 
whether any provision is made to insure a uniform number of kernels per acre 
in varietal tests in which varieties with different sized grains are included. 
Twenty-three replies were received. Eleven increase the rate of seeding for 
large-kerneled varieties. Of the remainder, io reported that they do not con- 
sider the variation in size important or else they consider it impractical to 
eliminate this source of error for small grains. 

(Did I 'intrusions of Xursery Plots. — Of 17 experimenters who reported 
on tlx i/' and dimensions of nursery plots for small grain, 13 use rows ap- 
proximately 1 rod in length. Four reported the use of rows greater in length 
' •''"!.• u ( row 100 i'itI 111 length or longer. One reports that 
blockl 16 I'M long by 5" inches wide are used and another reports the use of 
p!"> ' t''t vpiare. Those who use rows usually space them 1 foot apart, but 
this distance varu s from H inches to as much as 18 inches. 

I - '!:« (himbef of rows per plot, <; reported that single rows were used, 8 
ting Ol 11-111 two to four rows, and 3 used plots consisting of 
five rows for some or all tests. 

For pasture and hay crops, 4 reported the use of single row plots, 6 the use 

I J row »" 4-rou plots, and I the use of 10-rovv plots. One uses small blocks 



AGRONOMIC AFFAIRS. 



351 



Replication of Nursery Plantings.— For small grains, 4 indicated that they do 
not replicate their plantings, 5 duplicate their tests, 10 replicate from two to 
four times, and 6 replicate five times or more. For rowed crops, 5 do not 
replicate, 6 duplicate their tests, an equal number replicate from two to four 
times, and 2 replicate five times or more. 

Four use single plantings for hay and pasture crops, 5 duplicate, 4 use from 
three to five plantings, and 1 replicates five times or more. 

Of 23 who use or have used single rows for nursery plots only 7 make pro- 
vision for eliminating error due to the competition of adjoining rows. The 
impracticability of entirely eliminating such errors appears to have been one 
of the chief reasons for using plot's consisting of more than single rows. Prac- 
tically all who use single rows attempt to prevent erroneous conclusions by 
recording any unusual sources of error, as when some rows fail to germinate 
or are killed by cold or other causes. 

Cooperative Experiments with Farmers. — Of 30 who sent reports relating to 
cooperation with farmers all but 6 indicated that cooperative experiments of 
some kind were conducted. All but 1 investigator consider them necessary. A 
few experimenters were emphatic in their expression of the necessity of such 
tests, while others appeared to consider them of little value from an experi- 
mental viewpoint. 

Experiments in Technic. — Seven agronomists reported experiments, now 
under way relating to the technic of conducting field experiments with crops, 
and others have work planned for prosecution after the war. 

GENERAL DISCUSSION. 

It is fair to say that the results of this questionnaire show a very lively reali- 
zation of the importance of accurate field tests and an earnest endeavor to 
reduce such errors as far as possible with the money, time, and land available 
for experimental work. Many state that tests would be replicated a greater 
number of times and greater accuracy would be secured if it were possible to 
do so. It is especially encouraging to note that less than 20 percent of those 
who sent in reports depend on single plot's, but replicate their tests from one 
to several times. 

One need only compare the methods in general use by agronomists ten years 
ago with those used at the present time to appreciate the great improvement in 
technic that has taken place. That there is an appreciable variation in the soil 
from plot to plot was just beginning to be realized at that time. The use of 
check plots was by no means common and replication of plots was scarcely 
thought of. Variation of the soil is now one of the accepted tenets of agro- 
nomic science, and agronomists now replicate their tests almost as frequently 
as they formerly depended on single plots. 

Xo doubt much or all of this improvement in methods is due to the agitation 
which resulted in the appointment by the American Society of Agronomy of 
the first committee on the Standardization of Field Crop Experiments and to 
the excellent experiments relating to the technic of field experiments con- 
ducted by members of the Society. 

It would be surprising if the questionnaire had failed to show that there is 
need of more information of several kinds. There is much uncertainty as to 
the necessity of removing the border rows of field plots before harvest. There 



352 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



are those who contend that, as all plots are affected alike, the increase in yield 
due to the marginal effect is comparative and may be ignored, especially if the 
alleys are narrow. On the other hand, the fact that nearly half of those send- 
ing in reports remove the grain from the alleys at considerable expense and 
inconvenience at a very busy time of the year shows there are some who do 
not share this view. 

Much remains to be done before it can be safely asserted that any given 
size of plot is best for a given purpose, and experimenters are perhaps even 
more in the dark as to the number of replications necessary or desirable for 
each set of conditions. In nursery tests it is apparent that not all are con- 
vinced of the necessity of avoiding all chances of error due to the competition 
of adjoining rows. 

Conclusion. 

The committee regrets that the survey could not be made more complete. 
Questionnaires were directed to interested investigators at all of the State 
experiment stations, but only about 70 percent replied. The replies, however, 
include reports from all of the stations carrying on any considerable amount 
of field-plot experimental work and are sufficiently representative to give a fair 
idea of the various methods in operation. It will be seen from the foregoing 
summaries that there are great variations in the practices followed and many 
differences of opinion among the investigators concerned as regards the most 
desirable methods to follow. Circumstances, no doubt, alter cases and prob- 
ably it would not be wise to lay down any hard and fast rules from our present 
knowledge of the factors involved. The relative merits of different systems 
must receive further study and the committee strongly recommends that more 
of the members in position to do so undertake experimental studies with a view 
to determining the best and most practicable methods for the different lines of 
investigation requiring field-plot experiments. Doubtless a certain amount of 
standardization is possible and the members of this society should work together 
to that end. 

Additions to Bibliography. 

The following titles are to be added to the bibliography of the subject of 
field-plot experimentation published in the December, 1918, issue of this journal : 

AlJBCAlfDBOWITO H, [. Determinations of probable errors in field experiments. 

In Mitt. I h ut. Landw. Oesell., 28, no. 18, p. 268-271. 1913. 
Ai.way. P. J., and Kins, E. K. On the sampling of cultivated soils. In 25th 

Ann. Kpt. \ebr. Agr. Kxpt. Sta.. p. 52-55. 1912. 
Ai.way, K. J., and TfcUMBULL, R. S. On the sampling of prairie soils. In 

25th Ann. K'pt. Xcbr. Agr. Kxpt. Sta., p. 35-51. 1912. 
Ai.woon, W. I',., ;ui<! I'kki, R. II. Suggestions regarding size of plats. Va. 

Agr. Expt Sta. Bui. 6, 20 p. 1890. 
Babcoc k. I- k. Cereal experiment! at the Williston, N, Dak., substation. U.S. 

Dept. Agr. But 270. 191 5. 

V. (,. The experimental error in field tests. In Zhur. Opuitn. 
I n v Jour. Kxpt. Landw.), 17, no. 2, p. 99-121. 1916. Abs. in 

Expt. Sta. Rcc, 37, no. 6, p. 528. 191 7. 
BlLGtt, O. A method for variety tests. /;/ I litis. Landw. Ztg., 32, no. oj , p. 



AGRONOMIC AFFAIRS. 



353 



Cardon, P. V. Tillage and rotation experiments at Nephi, Utah. U. S. Dept. 

Agr. Bui. 157. 1015. 
Collins, G. N. A more accurate method of comparing first generation maize 

hybrids with their parents. In Jour. Agr. Research, 3, no. 1, p. 85-91. 1914. 
Davenport, E. Principles of Breeding, 721 p. Ginn & Co., Boston. 1907. 
, and Frazer, W. J. Experiments with wheat, 1888-1895. 111. Agr. Expt. 

Sta. Bui. 41, p. 153-155- 1896. 
De Vries, O. On the effect of using many parallel plats in field experiments. 

In Teysmannia, 26, no. 8-9, p. 465-474. 1915. 
Ehrenburg, P. The nitrogen economy of cultivated soils. In Fiihlings Landw. 

Ztg., 581, no. 1, p. 241-246. 1909. 
Fisher, Arne. The Mathematical Theory of Probabilities and its Application 

to Frequency Curves and Statistical Methods. The Macmillan Co., New 

York. 1916. 

v Gregoire, A. Field Experiments and the interpretation of their results. In 

Rpt. Tenth Cong. Internat. Agr. Gand., sec. 2, question 1, 13 p. 1913. 
Hartley, C. P., Brown, E. B., Kyle, C. H., and Zook, L. L. Crossbreeding 

corn. U. S. Dept. Agr., Bur. Plant Indus. Bui. 218, 66 p. 1912. 
Hays, H. K., and Arny, A. C. Experiments in field technic in rod row tests. 

In Jour. Agr. Research, 11, no. 9, p. 399-419. 1917. 
Hedrick, U. P. A comparison of tillage and sod mulch in an apple orchard. 

N. Y. State (Geneva) Agr. Expt. Sta. Bui. 314. 1909. 
. Is it necessary to fertilize an apple orchard? N. Y. State (Geneva) 

Agr. Expt. Sta. Bui. 339. 191 1. 
Hilgard, E. W. Soil tests and variety tests. In Proc. Soc. Prom. Agr. Sci., p. 

89-94. 1901. 

Hill, C. E. A drill for seeding nursery rows. In Jour. Amer. Soc. Agron., 
10: 165, 166. 1918. 

Hunnicutt, B. H. Some Brazilian problems in agronomy. In Jour. Amer. 

Soc. Agron., 5 : 34-38. 1913. 
Kiesselbach, T. A. Studies concerning the elimination of experimental error 

in comparative crop tests. In Nebr. Agr. Expt. Sta. Research Bui. 13, 

91 p. 1918. 

Klinck, L. S. The improvement of small grains at Macdonald college. In 

Proc. Amer. Soc. Agron., 4: 126-129. 1912. 
Kostecki, E. Methods of testing varieties. In Trudy Biuro Prikl. Bot. (Bui. 

Agnew. Bot.), 5, no. 7, p. 177-204. 1912. 
Lehn, D. New work on methods for variety testing. In Bl. Zuckerriibenbau, 

20, no. 3, p. 33-39; no. 4, P- 52-55- I9I3- 
Leidner, R. Field experiments and compensating calculations. In Landw. 

Jahrb., 49, no. 1, p. 105-135. 1916. Abs. in Expt. Sta. Record, 37, no. 6, 

p. 528. 1917- 

Love, H. H., and Craig, W. T. Methods and results obtained in cereal inves- 
tigations at the Cornell station. In Jour. Amer. Soc. Agron., 10: 145-157. 
1918. 

McCall, A. G. A new method for harvesting small grain and grass plot's. In 

Jour. Amer. Soc. Agron., 9: 1 38-141. 1917. 
McKee, Roland. Moisture as a factor of error in determining forage yields. 

In Jour. Amer. Soc. Agron., 6: 113-117. 1914. 



354 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Mivake. C. The experimental error in field trials and the effect on this error 
of various methods of sampling. In Ber. Ohara Inst. Landw. Forsch., I, 
no. i. p. 111-121. 1916. 

Moore. R. A. The testing of varieties as foundation work in the improve- 
ment of farm crops. In Proc. Amer. Soc. Agron., 1 : 27, 28. 1908. 

Xorgord. C. P. Crop demonstrations on State and County farms. Wis. Agr. 
Expt. Sta. Bui. 208. 191 1. 

Pearl. R.. and Miner, J. R. A table for estimating the significance of sta- 
tistical constants. Maine Agr. Expt. Sta. Bui. 226, p. 85-88. 1914. 

Rogalski, B. To what extent can the results of fertilizer and variety tests be 
influenced and thus lead to false conclusions? In Illus. Landw. Ztg., 34, 
no. 43, p. 400, 401 ; no. 4, p. 407, 408. 1914. 

Rose, R. E. Methods of conducting fertilizer tests. In Amer. Fert., 35, no. 10, 
p. 21-23. 101 1- 

Salmon, S. C. Cereal investigations on the Belle Fourche Experiment Farm. 
U. S. Dept. Agr. Bui. 297. 191 5. 

Schxeidewind, W. Experiments with different sized plats. In Mitt. Deut. 
Landw. Gesell., 39, no. 21, p. 298-300. 1914. 

Scholz, H. Methods of conducting variety trials. In Fiihlings Landw. Ztg., 
59, no. 22, p. 776-785 ; no. 23, p. 807-830. 1910. 

Schreiner, Oswald, and Skinner, J. J. The triangle system of fertilizer ex- 
periments. In Jour. Amer. Soc. Agron., 10: 225-246. 1918. 

Spragg, F. A. Elimination of error. In 28th Ann. Rpt. Mich. Agr. Expt. Sta., 
p. 226-228. 191 5. 

Stewart, John P. The fertilization of apple orchards. In Pa. Agr. Expt. Sta. 
Ann. Rpt. 1909-1910, p. 279-300. 1910. 

Ten Eyck, A. M. The testing of varieties as foundation work in the improve- 
ment of farm crops. In Proc. Amer. Soc. Agron., 1 : 33-39. 1909- 

V in all. H. X., and McKee, R. Moisture content and shrinkage of forage and 
relation of factors to the accuracy of experimental data. U. S. Dept. Agr. 
Bui. 353. 1916. 

von Rum KIR, K.. Lexdner, K. R., and Alexandrowirsch, I. The application of 
;i new method in variety tests of cereals. In Ztschr. Pflanzenzucht, 2, no. 2, 
p. 180-232. 1914. 

G I . Some business questions involved in the interpretation of fer- 
tili/< if tests. In Proc. Amer. Soc. Agron., 4: 62-66. 1912. 
WlAVCKO, A. T. The testing of varieties as foundation work in the improve- 
ment of farm (Tops. In Proc. Amer. Soc. Agron., 1 : 29-33. 1909. 

I) I it-Id cxpc-i in' hi Maine Agr. Expt. Sta. Bui. 224. 1914. 
7.AV1T/., C. A. Care and management of land used for experiments witli farm 
crops. In Proc. Amer. Soc. Agron., 4: 122-125. 1912. 

Kcspccl fully submitted, 

A. T. WlANCKO, 

F. S. Harris, 
S. C. Salmon, 

Committee, 



AGRONOMIC AFFAIRS. 



355 



REPORT OF THE COMMITTEE ON TERMINOLOGY. 

Your Committee on Terminology begs to report that, owing to the conditions 
arising out of the war, the work of the Committee has been practically sus- 
pended during the past year, but it is purposed now to continue it with vigor. 
Thus far four contributions to the subject have been published, and when the 
whole subject has been covered it is proposed that these be combined and pub- 
lished as an official glossary. Yours very truly, 

C. V. Piper, 
Chairman of the Committee. 

REPORT OF THE EDITOR. 

Because of the increased cost of publication and the reduced income of the 
Society due to loss of members to military service and for other reasons, it 
has been necessary to curtail the year's volume of the Journal of the Ameri- 
can Society of Agronomy. Including the December number, the volume con- 
sists of 360 pages, as compared with 432 pages printed in 1917 and 400 pages 
printed in 1916. Not including the December number, the forms for which are 
not yet made up, the 1918 volume has included 41 papers by 55 authors, repre- 
senting 19 States and the District of Columbia. This is one more paper than 
was included in the considerably larger volume of last year, showing a desirable 
tendency toward shorter papers. To illustrate these papers, 9 plates and 46 
text figures have been used. 

The editing of the Journal has been accomplished with considerable difficulty, 
as the editor has been either traveling or on emergency work outside Wash- 
ington during all except two months during the year. It has often been difficult 
to reach him promptly with papers for publication or with proof, and conse- 
quently the promptness and regularity of issuance of the Journal has not been 
up to the desired standard. As a large agricultural library has not usually been 
at hand, it has been impossible to check many of the citations of literature, and 
therefore these have not always been as accurate or as uniform as could be 
wished. The editor has refrained from resigning only because he has felt that 
others who might do the work are as heavily loaded with emergency duties 
and perhaps as badly handicapped as he. 

Necessarily, the size of the annual volume is dependent on the Society's 
income. It is hardly likely that printing costs will further increase, so that if 
the present membership can be maintained the 1919 volume should be at least 
as large as the one just published. With the favorable conclusion of the war, 
however, the Society's membership should be largely increased, and this in- 
crease will naturally bring prosperity to the Society's publication. The main- 
tenance or the progress of the Journal of the American Society of Agronomy 
is dependent largely on the whole-hearted support of the Society's membership. 

The editor regrets exceedingly that it is not possible for him to attend the 
eleventh annual meeting of the Society, the first that he has missed since his 
official connection with the organization. He trusts that it will be a very inter- 
esting and successful one, and that it will mark the beginning of a new era of 
progress for the American Society of Agronomy. 

Respectfully submitted, 

C. W. Warburton, 
Editor. 



INDEX. 



Page. 

Acidit\', soil, Litmus paper as a 

test for 180 

Acid soils, Effect of, on crops ... 45 

Address, Changes of, 

48, oj6, 144, 190, 223, 262, 308 

Address list of members 332 

Agronomic affairs, 

47. 94. 143. 189, 222, 262, 307, 326 

Albrecht, Win. A., paper on 
" Changes in the nitrogen 
content of stored soils "... 83 

Alfalfa, Glandular pubescence in . 159 

Aluminum, Influence of, on acid 

soils 45 

Alway, F. J., McDole, G. R., and 
Trumbull, R. S., paper on 
u Interpretation of field ob- 
servations on the moistness 
of the subsoil " 265 

Annual dues 94, 189 

Amy, A. C, see Garber, R. J. 

Bacterial activities of soil, Influ- 
ence of higher plants on ... 313 

Bailey, L H., see LeClerc, J. A. 

Ball, Carleton EL, and Clark, J. 
Allen, paper on 44 Naming 
wheat varieties " 87 

Bean, velvet, Origin of varieties of 175 

Bermuda khi^s seed. Germination 

of 279 

Biggar. II. Howard, paper on 
Primitive methods of maize 
seed preparation " 183 

Bizzcll. J. A., and Lyon, T. L., 
paper on "The effect of cer- 
tain factors on the carbon 
dioxide content of soil air" 
(Figi. 13-21) 97 

Boltz, George K„ paper on " Loss 
of organic matter in clover 
returned to the noil " 210 

Rothnakian, Sarkin, paper on "The 

% 



Page. 

mechanical factors deter- 
mining the shape of the 
wheat kernel" (Fig. 27) .. 205 

Breeding, Wheat 113 

Bryan, W. C, paper on 44 Hasten- 
ing the germination of Ber- 
muda grass seed by the sul- 
furic acid treatment" (PI. 9) 279 

Call, L. E., and Sewell, M. C, pa- 
per on 41 The relation of 
weed growth to nitric nitro- 
gen accumulation in the 
soil " 35 

Carbon dioxide content of soil air 97 

Cereal investigations at the Cor- 
nell station 145 

Cereals, Development of secondary 

rootlets in 32 

Ceretoma, Influence of, on nitro- 
gen-gathering functions of 
cowpea 256 

Changes of address, 

48, </>, 144, 190, 223, 262, 308 

Clark, J. Allen, see Ball, Carleton R. 

Clover returned to soil, Loss of 

organic matter in 210 

Coe, H. S., paper on 44 Origin of 
the Georgia and Alabama 
varieties of the velvet bean" 
(Figs. 25 and 26) 175 

< ommittee, Auditing 344 

Nominating 344 

Standardization of field ex- 
periments 344 

Terminology 354 

( onimittee memberships 189 

( ommittees, Reports of 344 

( orn, Preparation of seed 183 

K'elation between yield and 
e;ir characters in 250 

Self-fertilization in 123 

16 



INDEX. 



357 



Page. 

Cornell station, Cereal investiga- 
tions at 145 

Cotton, Time at which most moist- 
ure is used by 185 

Cowgill, H. B., paper on " Cross- 
pollination of sugar cane " . 302 

Cowpea, Influence of Ceretomaon 
nitrogen-gathering functions 
of 256 

Craig, W. T., see Love, H. H. 

Crop centers of the United States 49 

Cross-pollination, Natural in wheat 120 
of sugar cane 3° 2 

Cutthroat grass 162 

Davidson, J., and Le Clerc, J. A., 
paper on " The effect of so- 
dium nitrate applied at dif- 
ferent stages of growth on 
yield, composition, and qual- 
ity of wheat— 2" 193 

Davisson, B. S., and Sivaslian, G. 
K., paper on " The determi- 
nation of moisture in soils " 198 

Decomposition of organic matter 

in soils 281 

Development of secondary root- 
lets in cereals 3 2 

Drill for seeding nursery rows... 165 

Dunnewald, T. J., paper on "Vege- 
tation as an indicator of 
the fertility of sandy pine 
plains soils in northern Wis- 
consin " (Fig. 4) 19 

Ear characters of corn, Relation 

of yield to 250 

Editor, Report of the 351 

Effect of acid soils on crops, In- 
fluence of aluminum on . . . 45 

Einkorn, Milling and baking tests 

of : 215 

Emerson, Paul, paper on " A sim- 
ple method of demonstrat- 
ing the action of lime in 
soils" 158 

Emmer. Milling and baking tests 

of 215 



Page. 

Experiments, fertilizer, Triangle 

system for 225 

Fertilization, Self, in corn 123 

Fertilizer experiments, Triangle 

system for 225 

Field observations on moistness 

of subsoil, Interpretation of 265 

Food requirements, Mineral, of 

wheat at different stages... 127 

Freeman, Geo. F., paper on " A 
mechanical explanation of 
progressive changes in the 
proportions of hard and 
soft kernels in wheat " 23 

Funds collected by the treasurer. 330 

Gaines, E. F., paper on " Compar- 
ative smut resistance of 

Washington wheats " 218 

Garber, R. J., and Amy, A. C, pa- 
per on " Relation of size of 
sample to kernel percentage 
determinations in oats " ... 134 
Germination of Bermuda grass 

seed 279 

Glandular pubescence in Medicago 159 

Grass, Cutthroat 162 

Growth of sheep sorrel in cal- 
careous and dolomitic media 29 

Hartwell, Burt L., and Pember, 
F. R., paper on " Aluminum 
as a factor influencing the 
effect of acid soils on dif- 
ferent crops " 45 

Hayes, H. K., paper on " Natural 

cross-pollination in wheat " 120 
paper on " Normal self-fer- 
tilization in corn " 123 

Hendry, G. W., paper on " Rela- 
tive effects of sodium chlo- 
ride on the development of 
certain legumes " 246 

Hill, C. E., paper on " A drill for 

seeding nursery rows " 165 

Honor roll. . .95, 191, 223, 263, 308, 326 

Hutcheson, T. B., and Wolfe, T. 



35^ JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Page. 

K.. paper on " Relation be- 
tween yield and ear charac- 
ters in corn " 250 



Identification of oat varieties 



171 



Karraker, P. E., paper on " The 
value of blue litmus paper 
from various sources as a 
test for soil acidity" 180 

Kernel percentage determinations 

in oats 134 

Kernels, Proportions of hard and 

soft, in wheat 23 

Le Clerc, J. A., Bailey, L. H., and 
Wessling, H. L., paper on 
" Milling and baking tests 
of einkorn. emmer, spelt 

and Polish wheat" 215 

See also Davidson, J. 

Legumes, Effect of sodium chlo- 
ride on development of 246 

Leonard, Lewis T., and Turner, 
C. F., paper on " Influence 
of Ceretoma trifurcata on 
the nitrogen-gathering func- 
tions of the cowpea " 236 

Lime in soils, action of, Method 

of demonstrating 158 

Litmus paper as a test for soil 

acidity 180 

J <>f organic matter 210 

Love, ft IL, and Craig, W. T., 
papef OH " Methods used 
and results obtained in ce- 
real investigations at the 
Cornell station" (PL 4 and 

Fig. 24) 145 

Lyon. T. L.. paper on " Influence 
of higher plants on bacterial 
aetivihes in soils " (presi- 
dential address) 313 

See alio BifsaU, J. A. 

McCall, A. G., and Richards, P. 
B . paper on " Mineral food 
requirements of the wheat 



Page. 

plant at different stages of 
development (Pis. 2 and 3 
and Figs. 22 and 23) 127 

McClelland, C. K., paper on " The 
time at which cotton uses 
the most moisture" 185 

McDole, G. R., see Alway, F. J. 

Maclnt'ire, W. H., paper on " The 
growth of sheep sorrel in 
calcareous and dolomitic 
media " 29 

McKee, Roland, paper on " Glan- 
dular pubescence in various 
Medicago species" 159 

Maize seed preparation, Primitive 

methods of 183 

Manuscripts for publication, Prep- 
aration of 322 

Mechanical factors determining 
the shape of the wheat ker- 
nel 205 

Medicago, Glandular pubescence in 159 

Meeting of western agronomists. 311 

Members, Address list of 332 

Members deceased 144, 223 

lapsed 48, 143, 342 

New, 47, 95, 143, 190, 223, 262, 307 

reinstated 47, 190 

resigned, 48, 95, 143, 190, 223, 307 

Membership changes, 

47, 95, 143, 190, 222, 262, 307 

Membership, Secretary - treasur- 
er's report on 330 

Merkle, Fred. G., paper on "The 
decomposition of organic 
matter in soils" (Figs. 41-46) 281 

Method of demonstrating action 

of lime in soils 158 

Methods of maize seed prepara- 
tion. Primitive 183 

Methods used in cereal investiga- 
tions 145 

Milling tests of wheat 215 

Mineral food requirement! of 

wheat 127 

Minutes of the annual meeting... 343 



INDEX. 



359 



Page. 

Moistness of subsoil, Field obser- 
vations of 265 

Moisture, soil, Determination of.. 198 
Use of, by cotton 185 

Montgomery, E. G., paper on "The 
identification of varieties of 
oats in New York" 171 

Naming wheat varieties 87 

New members, 

47, 95, 143, 190, 223, 262, 307 
Nitrogen content of s.tored soils.. 83 
Nitrogen gathering functions of 

the cowpea 256 

Nitrogen, Nitric, accumulation in 

soil 35 j 

Nominating committee, Report 

of 344| 

Notes and news, 

48, 06, 144, 191, 223, 263, 309, 327 
Nursery rows, Drill for seeding. . 165 J 

Oats, in New York, Identification 

of varieties of 171 

Kernel percentage determina- 
tions in 134 

Officers elected for 1919 344 

Official changes 189 

Organic matter in soils, Decompo- 
sition of 281 

Loss of 210 

Pember, F. R., see Hartwell, B. L. 

Piper, C. V., paper on " Cutthroat 

grass " 162 

Pollination, Cross, in wheat 120 ; 

Polish wheat, Milling and baking 

tests of 215 : 

Preparation of manuscripts for 

publication 322 

Presidential address 313 

Relation of weed growth to nitric 

nitrogen accumulation 35 

Report of the editor 354 

of the secretary-treasurer. . . . 330 

Reports of committees 344 

Results of cereal investigations at 

the Cornell station 145 I 



Page. 

Richards,' P. E., see McCall, A. G. 

Rootlets, Secondary, in cereals . . 32 

Rye, Rosen 157 

Sample, Relation of size of, to 
kernel percentage determi- 
nations 134 

Schreiner, Oswald, and Skinner, 
J. J., paper on " The trian- 
gle system for fertilizer ex- 
periments " (Pis. 5-7 and 
Figs. 28-41) 225 

Secretary-treasurer, Report of . . . 330 

Seed tests, Variations in, resulting 

from sampling errors 1 

Sewell, M. C, see Call, L. E. 

Sivaslian, G. K., see Davisson, B. S. 

Skinner, J. J., see Schreiner, Oswald. 

Smith, L. H., see Walworth, E. H. 

Smut resistance of Washington 

wheats 218 

Snyder, Harry, paper on " Wheat 
breeding ideals " 113 

Sodium chloride, Effect of, on 

development of legumes . . . 246 

Sodium nitrate, Effect of, on 

wheat 193 

Soil acidity, Blue litmus paper as 

a test for 180 

Soil air, Carbon dioxide content of 97 

Soil bacteria, Influence of higher 

plants on 313 

Soil fertility, Vegetation as indi- 
cator of 19 

Soil moisture, Determination of.. 198 

Soils ; acid, Effect of, on crops ... 45 
Decomposition of organic 

matter in 281 

stored, Nitrogen content of . . 83 

Sorrel, sheep, Growth of, in cal- 
careous and dolomitic media 29 

Spelt, Milling and baking tests of 215 

Spragg, F. A., paper on " Red 

Rock wheat and Rosen rye" 167 

Standardization of experiments, 

Report of committee on.... 344 

Stevens, O. A., paper on " Varia- 
tion in seed tests resulting 



360 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Page. 



from errors in sampling " 

(Figs. 1-3) 1 

Subsoil, Interpretation of field ob- 
servations on moistness of . 265 
Sugar cane, Cross-pollination of.. 302 
Sulfuric acid treatment to hasten 

germination 279 

Terminology, Report of commit- 
tee on 354 

Triangle system for fertilizer ex- 
periments 225 

Trumbull, R. S., see Alway, F. J. 

Turne'r.C. F.,see Leonard, Lewis T. 

United States, Crop centers of the 49 

Variations in seed tests 1 

Varieties. Identification of oat . . . 171 

Xaming 89 

Vegetation as an indicator of soil 

fertility 19 

Velvet bean, Origin of varieties of 175 

Waller, Adolph E., paper on "Crop 
centers of the United States" 
(Figs. 5-12) 49 

Walworth, E. H., and Smith, L. 
H., paper on " Variations in 



Page. 

the development of sec- 
ondary rootlets in cereals". 32 
Warburton, C. W., paper on 
" The preparation of manu- 



scripts for publication" 322 

Weed growth, Relation of, to ni- 
trogen accumulation in soil 35 
Wessling, H. L., see LeClerc, J. A. 
Western agronomic conference .. 311 

Wheat breeding ideals 113 

Wheat, Cross-pollination in 120 

Wheat, Effect of sodium nitrate 

on ! 193 

kernel, Factors determining 

shape of the 205 

plant, Mineral food require- 
ments of 127 

Polish, Milling and baking 

tests of 215 

Proportion of hard and soft 

kernels in 23 

Red Rock 167 

varieties, Naming 89 

Washington, Smut resistance 
of 218 

Yield of corn as related to ear 

characters 250 



VOLUME 10 



NUMBER 9 



JOURNAL 

OF THE 

American Society of Agronomy 

DECEMBER, 1918 



CONTENTS 

Influence of Higher Plants on Bacterial Activities in Soils (Presidential Ad- 
dress). T. Lyttletox Lyon 313 

The Preparation of Manuscripts for Publication. C. W. Warburtox 322 

Agronomic Affairs. 

Delay in Publication of the Journal — The Year's Work — Honor Roll — Notes 

and News 326 

Report of the Secretary-Treasurer for 191 7 (Financial Statement— Address Dist 

of Members — Lapsed Members) 330 

Minutes of the Annual Meeting ." 343 

Reports of Committees (Standardization of Field Experiments — Terminology). 344 

Report of the Editor 354 

Index 356 

PUBLISHED BY THE SOCIETY 

41 NORTH QUEEN ST., LANCASTER, PA., 
and 

Washington, D. C. 



Issued January 15, 1919. 



Acceptance for mailing at special rate of postage provided for in section 1103, Act of 
October 3, 1917, authorized on June 29. 1918 



JOURNAL 



OF THE 



American Society of Agronomy 

Issued Monthly except in June, July, and August. 



Editor 
C. W. WARBURTON 

Associate Editors 
Crops: CHARLES V. PIPER 
Soils: T. LYTTLETON LYON 

Assistant Editors 

Crop Production, C. A. MOOERS Soil Physics, L. E. CALL 

Crop Breeding, L. H. SMITH Soil Chemistry, W. P. KELLEY 

Crop Chemistry, R. W. THATCHER Soil Biology, J. G. LIPMAN 



MANUSCRIPTS 

Suitable articles concerned with instruction, demonstration, experimentation or 
research in agronomy will be accepted for publication. It is understood that articles 
submitted for publication have not appeared previously elsewhere and that they will not 
be offered for simultaneous publication in other journals without the consent of the 

Editor of the Journal of the American Society of Agronomy. 

Papers of any length, between I page and 30 or 40 pages, can be used. Personal 

and institutional items of agronomic interest, suitable for inclusion in " Notes and 

News," are solicited. 

To be accepted for publication, manuscripts should be original typewritten copies 
1 not carbons) double- or triple-spaced, with wide margins. Special care should be 
(Oven to the proper indication of main heads and subheads in the text, to preparation 
.•■.nd descriptions of tables, to citations of literature and to illustrations. For fuller 
details see recommendations on page 28 of volume 3 of Proceedings and examples'ln- 
that and other volumes. 

All illustrations desired should accompany the manuscript, should be numbered 
ibed, and referred to in the text. Line drawings must be made in India ink 

and glossy vclox prints of photographs are preferred for half-tones. 

REPRINTS 

reprints of each article will be furnished free. Additional copies will be sup- 
it a nominal charge. Covers on same paper as the publication with printed title 
f o cover* $i.oo, and 1 cent for each additional copy. Orders for reprints andj 

overs should br *rnt to the Editor immediately on receipt of proof of the article. 



THE AMERICAN SOCIETY OF AGRONOMY 



OBJECT 

Article II. The object of the Society shall be the increase and dissemination oi 
knowledge concerning soils and crops and the conditions affecting them. 

MEMBERSHIP 

Article IV. Membership shall be of three kinds, active, associate and local. 
Active membership shall be limited to persons who are engaged in teaching agronomy 
or in scientific investigation in some branch of agronomy. Associate membership shall 
be composed of other persons interested in the object of the Society. Associate mem- 
bers shall be entitled to all the privileges of the Society except that of voting. Local 
members shall have no vote in the Society and shall not be entited to a copy of the 
printed proceedings without payment of an extra sum of money as provided in Article 
V of this Constitution. 

Active and associate membership may be secured by the endorsement in writing 
of some active member and upon approval by the President and Secretary and pay- 
ment of the annual dues. 

BY-LAWS 

1. The annual dues for each active and associate member shall be $2.50, and for 
each local member $.50, which are due and payable on January 1 of the year for which 
membership is held. 

2. Any member in arrears for dues for more than one year shall thereby forfeit 
membership, but may be restored to membership without action of the Society upon 
the payment of such arrears. 

Applications for membership should be sent to the Secretary-Treasurer, preferably 
accompanied by remittance for dues, to save correspondence. 

PUBLICATIONS 

Proceedings. Four volumes of Proceedings have been issued, as follows: 

Vol. 1, cloth, 238 pp., 39 papers, 1909. Vol. 3, cloth, 286 pp., 14 papers. 191 1. 
Vol. 2, cloth, 154 pp., 16 papers, 1910. Vol. 4. cloth, 160 pp., 20 papers, 1912. 

Journal (continuing the Proceedings) : 
Vol, 5, quarterly, paper, 256 pp., 1913- Vol. 7, bimonthly, paper, 320 pp., 191 5. 
Vol. 6, bimonthly, paper, 294 pp., 1914. Vol. 8, bimonthly, paper, 400 pp., 1916. 

Vols. 9 and 10, monthly except June, July, and August, paper, 432 pp., 1917 and 191& 

Price of Volumes 1 to 9, $2.00; Volume 10, $2.50; all postpaid. 

Single issues, VoL 5, 60 cents ; Vols. 6 to 8, 35 cents ; Vols. 9 and 10, 30 cents. 

Special reduced price to members for volumes 1 to 9, inclusive. 

Libraries and individuals are invited to place subscriptions for the current volume 
and orders for previous volumes with the Secretary-Treasurer, Lyman Carrier, 41 North 
Queen Street, Lancaster, Pa., or U. S. Department of Agriculture, Washington, D. C 



AMERICAN SOCIETY OF AGRONOMY 



OFFICERS 

President 

First Vice-President 

Second Vice-President 

Secretary-Treasurer 

COMMITTEES 

EXECUTIVE COMMITTEE 
Composed of the Officers of the Society 

COMMITTEE ON SOIL CLASSIFICATION AND MAPPING 
C. F. Marbut, chairman; F. J. Alway, E. O. Fippin. 

J. G. Mosier, C. A. Mooers. 

COMMITTEE ON STANDARDIZATION OF FIELD EXPERIMENTS 

A T. Wiancko, chairman; S. C. Salmon, F. S. Harris. 

COMMITTEE ON TERMINOLOGY 

Cbables V. Piper, chairman; Carleton R. Ball, ' H. L. Shantz. 

Consulting Members 
L. C. Corbett, O. F. Cook. 



T. L. Lyon 
A. G. McCall 
,C. B. Lipman 
.Lyman Carrier 



COMMITTEE ON VARIETAL NOMENCLATURE 
L G. Montgoiiiiy, chairman; H. K. Hayes, W. C. Etheridgf